I actually think challenges are very under-used. Better to throw it away on a low-percentage challenge than let the game expire with it unused. It's a bit like guessing and losing a game of Who Wants To Be A Millionaire despite having lifelines remaining.I get that you have a challenge to use but that doesn't mean you need to throw it away.
Love ya bro but you lost me at 60% success rate on a play that one quick replay, as well as the live shot, looked to be about as close to 0% as one can get. The defender fouled two players AND was clearly inside the circle….the only question was who they would send to the line.I actually think challenges are very under-used. Better to throw it away on a low-percentage challenge than let the game expire with it unused. It's a bit like guessing and losing a game of Who Wants To Be A Millionaire despite having lifelines remaining.
Let's suppose a coach has an average 60% success rate, and that they only use it in situations where it will save them some conceded free throws or get them an additional possession (on an out-of-bounds call or a block/charge or the like). So that's 60% of a possession (1.1 points) or 2x FTs saved (1.5 points), call it a 60% chance of 1.3 points or 0.65 points gained for using the challenge. Then compound it with the second challenge you get if you win, so a 60% chance of another 0.65 points - net-net, using your challenges is worth about a point in an NBA game, a pretty substantial amount. Sometimes you win the challenge(s) and sometimes you won't, of course. But if you let the game finish without using it, you have given up that point equity. It's a bit of a variation on the secretary problem - when do you use it, and how late in the game do you start lowering your standards on what call to use it for?
Now, most coaches are holding onto theirs into the 4th quarter, keeping the standard very high until they're more-or-less forced to lower their threshold later in the game - see this excellent article:
View attachment 77740
(I think that's challenges per game between the two coaches, so coaches are averaging about 1/2 a use per game each)
I'd argue this is a cognitive bias - clearly, there are more calls worth challenging in the first quarter than get challenged, and proportionally fewer in the 4th quarter, so it's something like the endowment effect going on. Bad calls have no distribution throughout the game, or at least, I can't conceive of an argument that they do (longtime hoopheads might argue about make-up calls and foul-balancing and such, but I'd say those largely aren't going to be challengeable anyway). To compound that issue, they should want to challenge early in the game! All else equal, a challenge in the 1st is better than a challenge in the 4th. Because if they do and they win it, they have more of the game's remainder to find another suitable situation to use their second challenge.
As that Sportico article says, there's an observable bias at play here:
"For the past four years, there has been a clear divide between coaches willing to use challenges early in games versus those preferring to save them for crunch time. In the 2022-23 season, for instance, Daigneault used 21 challenges prior to the fourth quarter while Mazzulla only used one. The Miami Heat’s Erik Spoelstra never challenged prior to the final period all last year. "
As noted by other wonks, the success rate this season of first-3-quarters challenges is much higher (58%) than 4th-quarter challenges (42%), with some of the latter clearly being born "by desperation, not merit". But also because of the use-it-or-lose-it aspect to them. Use them earlier, and desperation won't come into play nearly as much.
Not all challengeable situations are an equal throw of the dice, of course. We all know this - we've seen 100%-reverse-ers, and we've seen 50-50 calls where the star begged for the challenge so the coach did it, and we've seen no-hopers very late in a game where the coach is more or less hoping for a miracle reversal because they need it to preserve a shot to win. But I can assure you that along with every other Celtics fan, I have observed 60%+ type situations in the first and second quarters that go unchallenged. And I think all coaches are giving up a lot of outcome equity by not challenging more than they do already. You can see it in the above chart, frankly: they are letting 1st-half calls go by, and then settling for (almost certainly more poor) opportunities late in the game, when they think it matters more. Their threshold of bad-call-ness over which they challenge goes down as the game goes along.
But it doesn't matter more in the 4th, as anyone sensible knows. A point is a point regardless of game-situation leverage. Let the players and the pundits worry about "clutch play". If you can get an extra point for your team by being smarter about challenges than the rest of the league is, you do it*. There are no bonuses for doing it in higher-leverage situations.
Are other factors at play? Sure. But I'm not sure they really affect the calculus:
- If it's a 20-point blowout, "why bother" is a perfectly reasonable point of view. But if you can turn around a charge/block call earlier in the game, maybe it becomes a blowout earlier in the game, and you rest your starters more. I don't expect many challenges are done in 20-point games in the 4th quarter, Payton Pritchard demanding one in that early-season blowout notwithstanding.
- Hey, timeouts matter, why should we throw them away in risking a challenge we might not fully believe in, or think we need? That's fine, but most games end with teams (A) not using their discretionary timeout prior to getting crunched down to 2 TOs at 3 minutes remaining, so before then you've basically got a freebie, and (B) frankly most games end without coaches using both their last 2 closing-minutes TOs either. For a small fraction of one-possession games, you'll end up using both of those final timeouts, but even then, if you challenge while still possessing your freebie (the use-it-or-lose-it one), you're not risking anything that you're going to miss in the endgame.
- Challenges are also an opportunity for a timeout, to give the guys a breather and a pep talk. You want those when you want them, one might say - not necessarily just for the sake of challenging a call. I might need to stop a run, or change the defense, or draw up an ATO play to get things humming again. But on the other hand: If you win the first challenge, you've just gotten a free timeout. If your team is better than their opponents, you want to help them stay focused and having caught their breath as much as possible. And if you're a better coach, then you'd back yourself to improve the team's performance more with more timeout pep talks, right? So I'm not really seeing how this ends up a net downside.
- What if there's a late-game situation with an egregious, 100%-percenter call and I've used my challenges on lesser things earlier in the game? I mean, yes, we can construct scenarios like that. But the odds don't favor it - there's a reason the solution to the secretary problem isn't "just wait as long as possible". Your best-equity play is to stop earlier than emotions might make you want to. Second, challenging more and challenging earlier means slightly compromising your success rate, in theory, yes, but you also get a second challenge if you win it. That's point equity you don't want to just give up. And third, if it's close and late, you can persuade the refs to initiate a challenge themselves to review a close call - and you're more likely to get them to do so if you've already used and won your challenge(s) that evening.
How does CJM do with challenges? Pretty well! Arguably, among the best in the league. From that same article:
View attachment 77743
I would argue that the "Efficient frontier" of challenge volume is down-and-to-the-right of Joe Mazzulla today. Yes, it'd be great if he could make more challenges while maintaining his 77% success rate. But even if that rate goes down with higher volume, he's still helping the team by doing so. Because, again, if the game ends with him not having used them, particularly if it was a close game, it's something of a tactical tragedy - he's yielded points he could have gained. I rather doubt if CJM started challenging an extra 0.1 calls per game (to take him to Mike Brown territory), his rate would go down to Mike Brown's 40%. Maybe it goes to 70%, or even 60%, but that's still several extra reversed calls in a season, and any one of those could end up mattering to a game's outcome.
The same is doubly true for every other coach in the league, clearly. Until coaches are almost never ending games with an unused challenge, we'll have no idea what the sweet spot is between volume and success rate. But the math pretty clearly says they're too conservative right now - much like the math in the NFL said that coaches go for it on 4th down too little, a decade or so ago, and gradually they started going for it more. Same emotional situation: they want to avoid looking foolish / getting blamed for making a bet and then losing it. But that's not the best play to help the team win.
The Sportico article goes into a more-detailed calculation of win-probability-added through coaches' challenges, which is kinda amazing. Daigneault added a full win (100% WPA) over the course of the season through his challenges, Billups 118% WPA. Mazzulla got 77% of a win last year, about average. Some of these are luck-of-the-draw high-leverage challenges, like one Billups got with 11 seconds remaining in a one-possession game that turned a block into a charge, which added ~40% WPA just from that one call. But you can't depend on having those dramatic opportunities present themselves. Over the course of a season, the right play is clearly to challenge more and challenge earlier.
* and don't get me started about poor free-throw shooters who refuse to shoot underhanded out of vanity - that's a whole other topic, but the same argument applies. I'm more forgiving because players are I guess expected to be more emotional creatures whose motivation levels and self-respect and such are factors that are hard to measure... but imo for purely tactical issues like challenges, coaches should be cost-benefit-optimizing robots.
No one who watches the nba is giving the clippers anything more than the "theyre good but you know what's coming" critique. I'm not sure what your victory lap would be.I've got to start hitting the pavement now so I can be in shape in the spring when the Clippers flame out and the Bucks are pushing the Celtics to the brink. Going to be a lot of victory laps around the Port Cellar.
Yeah wow so many similarities for sure that I hadn’t even considered right down to quitting on their own team and teammates. At least Simmons was never voted out of playoff share money by his teammates. #rondodigAt every step of his career, he's tall Rondo. It remains an imperfect comp because Rondo did the best when the lights were brightest and Simmons is the polar opposite, but Rondo was an elite passer, excellent rebounder for his size, an elite defender when he wanted to be, and extremely unselfish. He was a poor outside shooter early in his career, and later he was afraid to go to the rim for fear of being fouled and embarrassed at the line.
I wonder whether there's an aspect of playing to the score that makes a won challenge early in the game far less valuable. E.g. you reverse an OOB call in the 1st quarter, and then your team gives it all back with lazy transition defense.I actually think challenges are very under-used. Better to throw it away on a low-percentage challenge than let the game expire with it unused. It's a bit like guessing and losing a game of Who Wants To Be A Millionaire despite having lifelines remaining.
Let's suppose a coach has an average 60% success rate, and that they only use it in situations where it will save them some conceded free throws or get them an additional possession (on an out-of-bounds call or a block/charge or the like). So that's 60% of a possession (1.1 points) or 2x FTs saved (1.5 points), call it a 60% chance of 1.3 points or 0.65 points gained for using the challenge. Then compound it with the second challenge you get if you win, so a 60% chance of another 0.65 points - net-net, using your challenges is worth about a point in an NBA game, a pretty substantial amount. Sometimes you win the challenge(s) and sometimes you won't, of course. But if you let the game finish without using it, you have given up that point equity. It's a bit of a variation on the secretary problem - when do you use it, and how late in the game do you start lowering your standards on what call to use it for?
Now, most coaches are holding onto theirs into the 4th quarter, keeping the standard very high until they're more-or-less forced to lower their threshold later in the game - see this excellent article:
View attachment 77740
(I think that's challenges per game between the two coaches, so coaches are averaging about 1/2 a use per game each)
I'd argue this is a cognitive bias - clearly, there are more calls worth challenging in the first quarter than get challenged, and proportionally fewer in the 4th quarter, so it's something like the endowment effect going on. Bad calls have no distribution throughout the game, or at least, I can't conceive of an argument that they do (longtime hoopheads might argue about make-up calls and foul-balancing and such, but I'd say those largely aren't going to be challengeable anyway). To compound that issue, they should want to challenge early in the game! All else equal, a challenge in the 1st is better than a challenge in the 4th. Because if they do and they win it, they have more of the game's remainder to find another suitable situation to use their second challenge.
As that Sportico article says, there's an observable bias at play here:
"For the past four years, there has been a clear divide between coaches willing to use challenges early in games versus those preferring to save them for crunch time. In the 2022-23 season, for instance, Daigneault used 21 challenges prior to the fourth quarter while Mazzulla only used one. The Miami Heat’s Erik Spoelstra never challenged prior to the final period all last year. "
As noted by other wonks, the success rate this season of first-3-quarters challenges is much higher (58%) than 4th-quarter challenges (42%), with some of the latter clearly being born "by desperation, not merit". But also because of the use-it-or-lose-it aspect to them. Use them earlier, and desperation won't come into play nearly as much.
Not all challengeable situations are an equal throw of the dice, of course. We all know this - we've seen 100%-reverse-ers, and we've seen 50-50 calls where the star begged for the challenge so the coach did it, and we've seen no-hopers very late in a game where the coach is more or less hoping for a miracle reversal because they need it to preserve a shot to win. But I can assure you that along with every other Celtics fan, I have observed 60%+ type situations in the first and second quarters that go unchallenged. And I think all coaches are giving up a lot of outcome equity by not challenging more than they do already. You can see it in the above chart, frankly: they are letting 1st-half calls go by, and then settling for (almost certainly more poor) opportunities late in the game, when they think it matters more. Their threshold of bad-call-ness over which they challenge goes down as the game goes along.
But it doesn't matter more in the 4th, as anyone sensible knows. A point is a point regardless of game-situation leverage. Let the players and the pundits worry about "clutch play". If you can get an extra point for your team by being smarter about challenges than the rest of the league is, you do it*. There are no bonuses for doing it in higher-leverage situations.
Are other factors at play? Sure. But I'm not sure they really affect the calculus:
- If it's a 20-point blowout, "why bother" is a perfectly reasonable point of view. But if you can turn around a charge/block call earlier in the game, maybe it becomes a blowout earlier in the game, and you rest your starters more. I don't expect many challenges are done in 20-point games in the 4th quarter, Payton Pritchard demanding one in that early-season blowout notwithstanding.
- Hey, timeouts matter, why should we throw them away in risking a challenge we might not fully believe in, or think we need? That's fine, but most games end with teams (A) not using their discretionary timeout prior to getting crunched down to 2 TOs at 3 minutes remaining, so before then you've basically got a freebie, and (B) frankly most games end without coaches using both their last 2 closing-minutes TOs either. For a small fraction of one-possession games, you'll end up using both of those final timeouts, but even then, if you challenge while still possessing your freebie (the use-it-or-lose-it one), you're not risking anything that you're going to miss in the endgame.
- Challenges are also an opportunity for a timeout, to give the guys a breather and a pep talk. You want those when you want them, one might say - not necessarily just for the sake of challenging a call. I might need to stop a run, or change the defense, or draw up an ATO play to get things humming again. But on the other hand: If you win the first challenge, you've just gotten a free timeout. If your team is better than their opponents, you want to help them stay focused and having caught their breath as much as possible. And if you're a better coach, then you'd back yourself to improve the team's performance more with more timeout pep talks, right? So I'm not really seeing how this ends up a net downside.
- What if there's a late-game situation with an egregious, 100%-percenter call and I've used my challenges on lesser things earlier in the game? I mean, yes, we can construct scenarios like that. But the odds don't favor it - there's a reason the solution to the secretary problem isn't "just wait as long as possible". Your best-equity play is to stop earlier than emotions might make you want to. Second, challenging more and challenging earlier means slightly compromising your success rate, in theory, yes, but you also get a second challenge if you win it. That's point equity you don't want to just give up. And third, if it's close and late, you can persuade the refs to initiate a challenge themselves to review a close call - and you're more likely to get them to do so if you've already used and won your challenge(s) that evening.
How does CJM do with challenges? Pretty well! Arguably, among the best in the league. From that same article:
View attachment 77743
I would argue that the "Efficient frontier" of challenge volume is down-and-to-the-right of Joe Mazzulla today. Yes, it'd be great if he could make more challenges while maintaining his 77% success rate. But even if that rate goes down with higher volume, he's still helping the team by doing so. Because, again, if the game ends with him not having used them, particularly if it was a close game, it's something of a tactical tragedy - he's yielded points he could have gained. I rather doubt if CJM started challenging an extra 0.1 calls per game (to take him to Mike Brown territory), his rate would go down to Mike Brown's 40%. Maybe it goes to 70%, or even 60%, but that's still several extra reversed calls in a season, and any one of those could end up mattering to a game's outcome.
The same is doubly true for every other coach in the league, clearly. Until coaches are almost never ending games with an unused challenge, we'll have no idea what the sweet spot is between volume and success rate. But the math pretty clearly says they're too conservative right now - much like the math in the NFL said that coaches go for it on 4th down too little, a decade or so ago, and gradually they started going for it more. Same emotional situation: they want to avoid looking foolish / getting blamed for making a bet and then losing it. But that's not the best play to help the team win.
The Sportico article goes into a more-detailed calculation of win-probability-added through coaches' challenges, which is kinda amazing. Daigneault added a full win (100% WPA) over the course of the season through his challenges, Billups 118% WPA. Mazzulla got 77% of a win last year, about average. Some of these are luck-of-the-draw high-leverage challenges, like one Billups got with 11 seconds remaining in a one-possession game that turned a block into a charge, which added ~40% WPA just from that one call. But you can't depend on having those dramatic opportunities present themselves. Over the course of a season, the right play is clearly to challenge more and challenge earlier.
* and don't get me started about poor free-throw shooters who refuse to shoot underhanded out of vanity - that's a whole other topic, but the same argument applies. I'm more forgiving because players are I guess expected to be more emotional creatures whose motivation levels and self-respect and such are factors that are hard to measure... but imo for purely tactical issues like challenges, coaches should be cost-benefit-optimizing robots.
If most of the difference between coaches comes down to whether you got a big call overturned late, doesn't thatThe Sportico article goes into a more-detailed calculation of win-probability-added through coaches' challenges, which is kinda amazing. Daigneault added a full win (100% WPA) over the course of the season through his challenges, Billups 118% WPA. Mazzulla got 77% of a win last year, about average. Some of these are luck-of-the-draw high-leverage challenges, like one Billups got with 11 seconds remaining in a one-possession game that turned a block into a charge, which added ~40% WPA just from that one call.
Yeah I wasn't talking about that one play in particular, but rather using it as a jumping-of point to rant about the general principle. Which is why I wrote an essay not a one-linerLove ya bro but you lost me at 60% success rate on a play that one quick replay, as well as the live shot, looked to be about as close to 0% as one can get. The defender fouled two players AND was clearly inside the circle….the only question was who they would send to the line.
I also don’t feel that many challenges can be overturned by the language of the rule….certainly nowhere close to 60% imo and it would still need to be on an impactful play and not only a side out which we saw twice in that Phoenix game.
Hmmm, but if Chauncey Billups' WPA was 118%, and 40% of that came from 1 (!) block/charge call with 11 seconds left, doesn't this imply that having no late-game challenge is a far, far bigger risk than ending the game with one in your pocket?...Which is what you get if the game ends and your challenge remains in your pocket.
Given that MIL has a glaring need at perimeter defense, seems like Griffin's strategy of giving him minutes hoping he'll be ready for the playoffs is better than Rivers' burying him on the bench so he won't be useful at all this year, but what do I know?I’ve been a huge Beauchamp guy since early last year. The kid can flat out play but it’s hard to fit in a 21-yr old w no high level experience in schemes at either end of the floor with veterans who understand how each players movement affects the entire set…on both ends. It sucks for him too with minites available around the league on younger teams.
Disaster probably started when Middleton got hurt - MIL has been pushing chips to the pot the entire time and didn't have much left at this point. Series of probably not great events/decisions since then. But maybe they can pull it together for a couple of months in the playoffs.I mean, most of us were pretty skeptical of the Bucks trading out Jrue for Lillard and the results speak for themselves. A real disaster in the making for that franchise.
Whenever you can acquire a coach with Doc's postseason track record over the past decade, you just get out of the way and let him cook. Don't disturb the art.Given that MIL has a glaring need at perimeter defense, seems like Griffin's strategy of giving him minutes hoping he'll be ready for the playoffs is better than Rivers' burying him on the bench so he won't be useful at all this year, but what do I know?
That's fair, as long as you don't take said victory laps if the Clippers flame out due to injury. We've all caveated that they're an "if healthy" team enough, I think.I've got to start hitting the pavement now so I can be in shape in the spring when the Clippers flame out and the Bucks are pushing the Celtics to the brink. Going to be a lot of victory laps around the Port Cellar.
What is tricky about success rates, and I’ll point specifically to last nights Suns-Bucks game since that was the one most watched last night. There was a 1H foul whistled on the floor that was certain to get overturned but it was a side out and even Vogel had this snicker like yeah that’s the worst call ever but it isn’t optimal with the play not affecting the scoreboard or putting his guy in foul trouble. Then later in the game we of course had the Doc one that was used as a true flier with little chance to succeed BUT as you say you have one in your pocket so take a shot (even though that play had no shot).Yeah I wasn't talking about that one play in particular, but rather using it as a jumping-of point to rant about the general principle. Which is why I wrote an essay not a one-liner
The success rate thus far this year is 52%. On challenged out-of-bounds calls, it's 72% (foul calls do much worse at getting reversed). There are more suitable opportunities in a game than are currently used by coaches. Maybe that one wasn't it, but if so I'd argue that there were probably good-enough opportunities earlier in the game. And here's the main thing: even if it's 50-50, taking that 50% shot is still better than nothing! Which is what you get if the game ends and your challenge remains in your pocket.
My latest hot take is that after spending months thinking about the Sixer and Bucks, the Celts won't play either of them in the playoffs.I've got to start hitting the pavement now so I can be in shape in the spring when the Clippers flame out and the Bucks are pushing the Celtics to the brink. Going to be a lot of victory laps around the Port Cellar.
Orlando -> Indiana -> Miami would not be a shocking path for a C's playoff run.My latest hot take is that after spending months thinking about the Sixer and Bucks, the Celts won't play either of them in the playoffs.
Hey I was the one early in the year saying Beauchamp wasn’t playing enough so you’re preaching to the choir on this one. I do recognize that it’s a challenge to get young players into the rotation of a veteran team however.Given that MIL has a glaring need at perimeter defense, seems like Griffin's strategy of giving him minutes hoping he'll be ready for the playoffs is better than Rivers' burying him on the bench so he won't be useful at all this year, but what do I know?
Disaster probably started when Middleton got hurt - MIL has been pushing chips to the pot the entire time and didn't have much left at this point. Series of probably not great events/decisions since then. But maybe they can pull it together for a couple of months in the playoffs.
Yeah, the desire to turn a whole set of nuanced comment into a simplistic "Clippers great/Clippers a joke" binary choice would even embarrass WEEI callers.No one who watches the nba is giving the clippers anything more than the "theyre good but you know what's coming" critique. I'm not sure what your victory lap would be.
Interesting analysis. I do think Win Probability Added is probably the right prism to view this through, and speaks to why so many coaches are relatively conservative with their challenges. Even successfully challenging and overturning a 3 shot foul in the 1st quarter will have a neglibile impact on WPA, whereas like you showed critical challenges late can have huge swings. As a coach you want your challenges available to be able to take the huge swings late, but you never know if the opportunity will present itself (and frankly it very rarely does). Optimal challenge usage is definitely higher then the current levels like you said; coaches will always be afraid of not having a challenge to overturn a bad all late though.I actually think challenges are very under-used. Better to throw it away on a low-percentage challenge than let the game expire with it unused. It's a bit like guessing and losing a game of Who Wants To Be A Millionaire despite having lifelines remaining.
Let's suppose a coach has an average 60% success rate, and that they only use it in situations where it will save them some conceded free throws or get them an additional possession (on an out-of-bounds call or a block/charge or the like). So that's 60% of a possession (1.1 points) or 2x FTs saved (1.5 points), call it a 60% chance of 1.3 points or 0.65 points gained for using the challenge. Then compound it with the second challenge you get if you win, so a 60% chance of another 0.65 points - net-net, using your challenges is worth about a point in an NBA game, a pretty substantial amount. Sometimes you win the challenge(s) and sometimes you won't, of course. But if you let the game finish without using it, you have given up that point equity. It's a bit of a variation on the secretary problem - when do you use it, and how late in the game do you start lowering your standards on what call to use it for?
Now, most coaches are holding onto theirs into the 4th quarter, keeping the standard very high until they're more-or-less forced to lower their threshold later in the game - see this excellent article:
View attachment 77740
(I think that's challenges per game between the two coaches, so coaches are averaging about 1/2 a use per game each)
I'd argue this is a cognitive bias - clearly, there are more calls worth challenging in the first quarter than get challenged, and proportionally fewer in the 4th quarter, so it's something like the endowment effect going on. Bad calls have no distribution throughout the game, or at least, I can't conceive of an argument that they do (longtime hoopheads might argue about make-up calls and foul-balancing and such, but I'd say those largely aren't going to be challengeable anyway). To compound that issue, they should want to challenge early in the game! All else equal, a challenge in the 1st is better than a challenge in the 4th. Because if they do and they win it, they have more of the game's remainder to find another suitable situation to use their second challenge.
As that Sportico article says, there's an observable bias at play here:
"For the past four years, there has been a clear divide between coaches willing to use challenges early in games versus those preferring to save them for crunch time. In the 2022-23 season, for instance, Daigneault used 21 challenges prior to the fourth quarter while Mazzulla only used one. The Miami Heat’s Erik Spoelstra never challenged prior to the final period all last year. "
As noted by other wonks, the success rate this season of first-3-quarters challenges is much higher (58%) than 4th-quarter challenges (42%), with some of the latter clearly being born "by desperation, not merit". But also because of the use-it-or-lose-it aspect to them. Use them earlier, and desperation won't come into play nearly as much.
Not all challengeable situations are an equal throw of the dice, of course. We all know this - we've seen 100%-reverse-ers, and we've seen 50-50 calls where the star begged for the challenge so the coach did it, and we've seen no-hopers very late in a game where the coach is more or less hoping for a miracle reversal because they need it to preserve a shot to win. But I can assure you that along with every other Celtics fan, I have observed 60%+ type situations in the first and second quarters that go unchallenged. And I think all coaches are giving up a lot of outcome equity by not challenging more than they do already. You can see it in the above chart, frankly: they are letting 1st-half calls go by, and then settling for (almost certainly more poor) opportunities late in the game, when they think it matters more. Their threshold of bad-call-ness over which they challenge goes down as the game goes along.
But it doesn't matter more in the 4th, as anyone sensible knows. A point is a point regardless of game-situation leverage. Let the players and the pundits worry about "clutch play". If you can get an extra point for your team by being smarter about challenges than the rest of the league is, you do it*. There are no bonuses for doing it in higher-leverage situations.
Are other factors at play? Sure. But I'm not sure they really affect the calculus:
- If it's a 20-point blowout, "why bother" is a perfectly reasonable point of view. But if you can turn around a charge/block call earlier in the game, maybe it becomes a blowout earlier in the game, and you rest your starters more. I don't expect many challenges are done in 20-point games in the 4th quarter, Payton Pritchard demanding one in that early-season blowout notwithstanding.
- Hey, timeouts matter, why should we throw them away in risking a challenge we might not fully believe in, or think we need? That's fine, but most games end with teams (A) not using their discretionary timeout prior to getting crunched down to 2 TOs at 3 minutes remaining, so before then you've basically got a freebie, and (B) frankly most games end without coaches using both their last 2 closing-minutes TOs either. For a small fraction of one-possession games, you'll end up using both of those final timeouts, but even then, if you challenge while still possessing your freebie (the use-it-or-lose-it one), you're not risking anything that you're going to miss in the endgame.
- Challenges are also an opportunity for a timeout, to give the guys a breather and a pep talk. You want those when you want them, one might say - not necessarily just for the sake of challenging a call. I might need to stop a run, or change the defense, or draw up an ATO play to get things humming again. But on the other hand: If you win the first challenge, you've just gotten a free timeout. If your team is better than their opponents, you want to help them stay focused and having caught their breath as much as possible. And if you're a better coach, then you'd back yourself to improve the team's performance more with more timeout pep talks, right? So I'm not really seeing how this ends up a net downside.
- What if there's a late-game situation with an egregious, 100%-percenter call and I've used my challenges on lesser things earlier in the game? I mean, yes, we can construct scenarios like that. But the odds don't favor it - there's a reason the solution to the secretary problem isn't "just wait as long as possible". Your best-equity play is to stop earlier than emotions might make you want to. Second, challenging more and challenging earlier means slightly compromising your success rate, in theory, yes, but you also get a second challenge if you win it. That's point equity you don't want to just give up. And third, if it's close and late, you can persuade the refs to initiate a challenge themselves to review a close call - and you're more likely to get them to do so if you've already used and won your challenge(s) that evening.
How does CJM do with challenges? Pretty well! Arguably, among the best in the league. From that same article:
View attachment 77743
I would argue that the "Efficient frontier" of challenge volume is down-and-to-the-right of Joe Mazzulla today. Yes, it'd be great if he could make more challenges while maintaining his 77% success rate. But even if that rate goes down with higher volume, he's still helping the team by doing so. Because, again, if the game ends with him not having used them, particularly if it was a close game, it's something of a tactical tragedy - he's yielded points he could have gained. I rather doubt if CJM started challenging an extra 0.1 calls per game (to take him to Mike Brown territory), his rate would go down to Mike Brown's 40%. Maybe it goes to 70%, or even 60%, but that's still several extra reversed calls in a season, and any one of those could end up mattering to a game's outcome.
The same is doubly true for every other coach in the league, clearly. Until coaches are almost never ending games with an unused challenge, we'll have no idea what the sweet spot is between volume and success rate. But the math pretty clearly says they're too conservative right now - much like the math in the NFL said that coaches go for it on 4th down too little, a decade or so ago, and gradually they started going for it more. Same emotional situation: they want to avoid looking foolish / getting blamed for making a bet and then losing it. But that's not the best play to help the team win.
The Sportico article goes into a more-detailed calculation of win-probability-added through coaches' challenges, which is kinda amazing. Daigneault added a full win (100% WPA) over the course of the season through his challenges, Billups 118% WPA. Mazzulla got 77% of a win last year, about average. Some of these are luck-of-the-draw high-leverage challenges, like one Billups got with 11 seconds remaining in a one-possession game that turned a block into a charge, which added ~40% WPA just from that one call. But you can't depend on having those dramatic opportunities present themselves. Over the course of a season, the right play is clearly to challenge more and challenge earlier.
* and don't get me started about poor free-throw shooters who refuse to shoot underhanded out of vanity - that's a whole other topic, but the same argument applies. I'm more forgiving because players are I guess expected to be more emotional creatures whose motivation levels and self-respect and such are factors that are hard to measure... but imo for purely tactical issues like challenges, coaches should be cost-benefit-optimizing robots.
Challenges late in the game can be higher-leverage, yes. But again, not all will be successful, and it's better to end the game having shot your shot than having pocketed it. And as the stats show, if you wait until the 4th, the perfect opportunity may not come and you have likely already overlooked your best chances to get something reversed. Or two of them, even.I wonder whether there's an aspect of playing to the score that makes a won challenge early in the game far less valuable. E.g. you reverse an OOB call in the 1st quarter, and then your team gives it all back with lazy transition defense.
Obviously you can't take that logic too far, and the whole game does matter, but I wouldn't be surprised if a successful challenge late in the game is more valuable, all else being equal.
The Celtics clearly also feel that you don't want to challenge judgement calls early in the game, and try to save those challenges only for OOP-type, very objective calls. 3-point fouls also seem like valuable challenge territory.
No, you're confusing anecdote / outlier with data / averages. We can pick out individual situations that make dramatic differences - like the one Billups call - but we can then turn around and note that over a multi-year track record, Billups is worse at winning calls than Mazzulla (or, indeed, than league average), and has also added less WPA. Billups getting 40% WPA because a single dramatic chance fell his way does not change the long-term averages as expressed in season(s)-long success rate.If most of the difference between coaches comes down to whether you got a big call overturned late, doesn't that
* validate the idea that late-game overturns are more valuable
* invalidate most of the Sportico analysis?
I'm pretty uncomfortable saying Mazzulla is better than coaches further down on the list, or worse than Billups, just because Billups had the opportunity to make a block/charge challenge with 11 seconds left.
It's all interesting, so thanks for that.
I will be setting the parameters for my own victory lap, thank you very much.That's fair, as long as you don't take said victory laps if the Clippers flame out due to injury. We've all caveated that they're an "if healthy" team enough, I think.
I'd very pretty surprised if the Bucks didn't face the Celtics, I think they make the ECF (or perhaps play the Celtics in an earlier round). The Sixers, especially given the Embiid news, could totally flatline between now and the end of the regular season. I don't think that is even a controversial take at this point--they lost their best player for an extended period of time and we don't know how healthy he is going to be come playoff time.My latest hot take is that after spending months thinking about the Sixer and Bucks, the Celts won't play either of them in the playoffs.
Have you been watching the Bucks lately.. like any of their games? Because you really do love congratulating yourself for seeing things nobody else is seeing right now. If the Bucks go out in the first round.. what will you do then?I've got to start hitting the pavement now so I can be in shape in the spring when the Clippers flame out and the Bucks are pushing the Celtics to the brink. Going to be a lot of victory laps around the Port Cellar.
Yes, if a bad foul call doesn't result in points, then given that foul calls are overturned at a much lower rate (being more subjective), it's probably not the best opportunity to use it. And most NBA "bad calls" are not really reversable / reviewable either - things like travels, moving-screens, or especially whether contact on a play at the rim was incidental or a foul. So there are not a million calls per game that could be effectively (let's say 40%+ chance of reversal) challenged by coaches. Some games may even have the entire game elapse without one presentable opportunity, and the coaches sorta had no choice but to pocket their challenges. But I think there's pretty clearly more opportunities than what they're going for today.What is tricky about success rates, and I’ll point specifically to last nights Suns-Bucks game since that was the one most watched last night. There was a 1H foul whistled on the floor that was certain to get overturned but it was a side out and even Vogel had this snicker like yeah that’s the worst call ever but it isn’t optimal with the play not affecting the scoreboard or putting his guy in foul trouble. Then later in the game we of course had the Doc one that was used as a true flier with little chance to succeed BUT as you say you have one in your pocket so take a shot (even though that play had no shot).
Right here you have two plays, that are common over the course of most NBA games, that skew the “success rate” of challenges due to the timing and lack of impact of the bad call.
Admit that I was wrong. I don't have the type of pride where I'm going to be worried about looking incorrect here.Have you been watching the Bucks lately.. like any of their games? Because you really do love congratulating yourself for seeing things nobody else is seeing right now. If the Bucks go out in the first round.. what will you do then?
I will fight to my death on this one though. Points early in a game are FAR less valuable than down the stretch even though they are all counted equally in the scorebook. The 1H and really first 3Q (and often times more) are played in rhythm without much attention payed to the score of the game. In the final 3-6 min or so the games are managed by the score especially when you get down to the final minute. Those 1-2 points early in the game aren’t going to have nearly the same amount of impact as they will in the final possession due to both coaches playing to the score.An overturn is worth the same number of expected points whether it happens early or late game.
Just because the late-game ones can be high-leverage as to game outcome doesn't mean they're more valuable, it just means that that play swung things more in
Seriously. People love the strawman "clippers are title contenders!" bullshit just so they can dump on them in the playoffs. It's an annual, ignorant, nba tradition. I hate it. It's lazy af and has been going on forever.That's fair, as long as you don't take said victory laps if the Clippers flame out due to injury. We've all caveated that they're an "if healthy" team enough, I think.
How does the trade deadline play into this? Because the bucks wont be the same team after February 8.Admit that I was wrong. I don't have the type of pride where I'm going to be worried about looking incorrect here.
I simply think many posters here have been too dismissive of the Bucks chances of success in the postseason, and I've made that clear many times. If I'm proven right, there is going to be some victory laps.
It's a nuanced point, but I understand that, in point expectation, the reversed calls have the same value regardless of the point in the game.Just because the late-game ones can be high-leverage as to game outcome doesn't mean they're more valuable, it just means that that play swung things more in the moment, because there's less remaining game for the teams' skill differences to prove out (if any). A reversed call is, more or less, worth the same whether it comes early or late, in terms of points. This is a hard distinction to grasp, I'm having a hard time getting the right words for it. Think of this way: Suppose that one time per game, a manager in MLB can flip a coin to add a run for his team. If he waits to flip that coin until the 9th inning, it's more likely to end up affecting the game. But it's still worth the same to the team even if he flips it earlier in the game. It might change his endgame tactics (e.g. use a closer vs not) if he flips earlier, and if he waits he's more likely to know whether it'll matter (a blowout vs a close one). But after 9 innings, his team's runs are almost certainly going to be the same whether he flipped the coin early or late. The higher leverage with flipping it late doesn't mean it's better to wait. It's the same expected-value either way.
Look you guys just keep wanting to bury yourselves deeper and deeper, that is fine by me.How does the trade deadline play into this? Because the bucks wont be the same team after February 8.
However, the team they have today is not built to win in the playoffs.
Well, the "Locker Room" more or less ran Coach Bud out of town, and MIL made the Jrue for Dame move because the "Locker Room" wouldn't sign an extension until they "upgraded" plus the "Locker Room" didn't want Nurse, who has a ring last I checked and the "Locker Room" won't set picks for Dame in any meaningful way.From what we hear now it seems pretty clear that Griffin lost that locker room and a chance had to be made. There was a reason he’d been passed over in previous years for head jobs and it looks like those GMs were correct.
Like.Whenever you can acquire a coach with Doc's postseason track record over the past decade, you just get out of the way and let him cook. Don't disturb the art.
Wait. What? What do you think the Bucks are going to do today/tomorrow? I thought the general consensus is that they really don't have much to work with. I guess I've seen Portis for Wiggins, but that doesn't seem to move the needle unless Wiggins has some kind of complete turnaround with a new home.How does the trade deadline play into this? Because the bucks wont be the same team after February 8.
However, the team they have today is not built to win in the playoffs.
Honestly, I feel bad for the way Bud got run out. He wins a title, then loses a tough series to the Celtics without Middleton. 2023 hits, his brother shockingly dies, Giannis was hurt, they lose to the Heat and then suddenly he's gone as coach because Giannis (who has since proven himself to be an awful "GM") decides that's what he wants.Like.
Going from Bud's post-season record to Doc's post-season record, even if not intentional, is still, errr, puzzling.
Isn't "disaster" a bit strong? I was skeptical at the time, and the skepticism has been rewarded. But I could see the merit in rolling the dice and not standing pat. Jrue Holiday isn't the #2 offensive option overall / #1 crunch time offense option on a title-winning team, which is where he found himself last year with the Heat series on the line and Middleton compromised. Would they be better off having not made the trade? Well... maybe a little bit, but that team wasn't going anywhere as constructed anyway and Giannis was making noises about wanting out. I think "swing big" was the right idea even if the odds were against it working out.I mean, most of us were pretty skeptical of the Bucks trading out Jrue for Lillard and the results speak for themselves. A real disaster in the making for that franchise.
I don't know what they're going to do but they won't stand pat. They'll try something. They started with dame, then doc. When franchise start flailing, it doesnt stop until they get bounced in the playoffs.Wait. What? What do you think the Bucks are going to do today/tomorrow? I thought the general consensus is that they really don't have much to work with. I guess I've seen Portis for Wiggins, but that doesn't seem to move the needle unless Wiggins has some kind of complete turnaround with a new home.
Maybe they'll hire a new coachI don't know what they're going to do but they won't stand pat. They'll try something.
I think the "Brad move" in this spot would have been to try and address the root issue (lack of scoring creation) while not trading Jrue. The Bucks had some OK draft capital to work with, and you can go out then and look for top-50 type players who can score the ball, while still keeping the Giannis/Jrue core. Trading Middleton would have made sense too, although I think that would have had to wait until Dec.Isn't "disaster" a bit strong? I was skeptical at the time, and the skepticism has been rewarded. But I could see the merit in rolling the dice and not standing pat. Jrue Holiday isn't the #2 offensive option overall / #1 crunch time offense option on a title-winning team, which is where he found himself last year with the Heat series on the line and Middleton compromised. Would they be better off having not made the trade? Well... maybe a little bit, but that team wasn't going anywhere as constructed anyway and Giannis was making noises about wanting out. I think "swing big" was the right idea even if the odds were against it working out.
Now: getting a better coach definitely would have helped. That part _was_ a disaster.
Yeah but most of that is Giannis, no?On the Bucks front...they don't look great now, and remains to be seen if they can do anything to upgrade their bench before tomorrow,. But the 4 man lineup of Lopez-Lillard-Middleton-Giannis has been crushing (16.4 net rating in almost 600 minutes, just barely ahead of BOS's best 4 man combo w/ extended minutes of Tatum-Brown-White-Porzingis at +16.1). Health issues with Middleton nonwithstanding, I wouldn't necessarily be jumping at the chance to play them in round 2 with that group buoying them.
At least Bud got a title, which they can never take away (although, as Brian Billick once said, they won't stop trying).Honestly, I feel bad for the way Bud got run out. He wins a title, then loses a tough series to the Celtics without Middleton. 2023 hits, his brother shockingly dies, Giannis was hurt, they lose to the Heat and then suddenly he's gone as coach because Giannis (who has since proven himself to be an awful "GM") decides that's what he wants.
Seems like an obvious hire for a "trying to establish a winning culture" young team, like the pre-Ime Rockets.I wonder if Bud will coach again.
Also, for a small market team, remember that the biggest "win" in the Dame trade was the extension Giannis signed not long after. That to me justifies the trade even if the on-court results have been disappointing.Disaster is way too strong. Through 51 games they are 33-18. Last year they were 34-17. 31-20 the previous year. 32-19 the year they won the title
Good post. Yes, the challenge is unusual in the non-renewable nature of it, and makes it a more complicated thing to evaluate (hence my reference to formal mathematics thought exercises like the Secretary Problem). And yes, you don't know if you're going to be in a blowout or a close game. But:It's a nuanced point, but I understand that, in point expectation, the reversed calls have the same value regardless of the point in the game.
To state what I'm thinking about more formally:
* NBA games have a wide range out outcomes
* many games are decided by wide margins (much more than a possession), in either direction
* you don't know until the last couple minutes which "type" of game you're in
* all is being equal (I know, it never is), being able to choose to challenge when you find yourself in one of the close games is an advantage
The challenge is pretty unique, relative to other actions a team can take to improve expected point outcomes during the game, because it can be burned. If you shoot an open 3 and miss, you don't lose the ability to shoot 3s in the last 2 minutes. Ditto for any other game action.
For a team like the Celtics, whose point differential is such that the expected game is often not close, I wonder whether there is reduced value in winning early challenges, because you know going in that you're less likely to be in a close one at the end. Once you find out that you're in one, having a challenge in hand, even with a lower probability of success if/when you use it, could be more valuable.
I might be talking out of my ass, but I think that the "use it and maybe lose it" nature of challenges adds more to the analysis than you can get from a simple expected points calculation.
Every forum needs its Kevin Durant chaos-agent. Keeps things fun, as you said. Just gotta be able to keep track of what you actually believe, or it can devolve quickly into just common trolling.Look you guys just keep wanting to bury yourselves deeper and deeper, that is fine by me.
Just so everyone knows, I'm leaning into a bit here and none of this personal. I think it WILL be fun this spring, either I get to do a victory lap, or everyone dunks on me.
There is a difference between leverage and value which is maybe what you are getting at here, but the correct way to think about the challenge is maximizing value and not leverage.I will fight to my death on this one though. Points early in a game are FAR less valuable than down the stretch even though they are all counted equally in the scorebook.
Well, Brad isn't operating in a tiny market with an unhappy superstar threatening to leave.I think the "Brad move" in this spot would have been to try and address the root issue (lack of scoring creation) while not trading Jrue. The Bucks had some OK draft capital to work with, and you can go out then and look for top-50 type players who can score the ball, while still keeping the Giannis/Jrue core. Trading Middleton would have made sense too, although I think that would have had to wait until Dec.
The issue with the Dame approach is that they lost Jrue for a marginal upgrade and spent the draft capital, which just doesn't leave much room to maneuver at that point.
Id like to expand on this point because it trips people up a lot. To illustrate lets imagine a very simple game. You get 3 coin flips, and get 1 point if you draw a heads and -1 point if you get tails. The goal is to end the game with > 0 points. You are allowed one "challenge" to a flip, where you can modify a tails to a heads.There is a difference between leverage and value which is maybe what you are getting at here, but the correct way to think about the challenge is maximizing value and not leverage.
Just glancing at the list, I think he underrates Chet and overrates Banchero and Jaylin Williams. Williams in particular being above Chet I don't understand, and that includes when Williams was getting some ASG buzz earlier. Williams is a very good player, but Chet is younger, pretty much just as good offensively (and more versatile) and a significantly better defender, and will likely be one of the 3-4 best defensive players in the NBA in his prime if he can add even a little more weight.Simmons made an updated trade value list and video:
https://nbarankings.theringer.com/
View: https://youtu.be/QDmx7ej9IVM?si=ae0hnZkfO2QuSoPC
Celtics: Holiday at 63, Porzingis at 56, White at 47, Brown at 38, Tatum at 6.
The main thing I was curious about with this list was how many guys were above Wemby. Just three (Jokic, Giannis and Luka), with four just behind in the same category (SGA, Tatum, Edwards, Banchero). I wonder where Embiid would've ranked if he'd pulled this together two weeks ago, right after the 70 point game (he was 12th on this list).
I think injury and durability concerns come into play here, which is going to be a problem for Chet his entire career (unless he starts ripping off multiple 70+ game seasons, at least). I've always loved Banchero (I've posted here that he was the best player drafted in the 4+ years between Luka and Wemby), I'm fine with his placement.Just glancing at the list, I think he underrates Chet and overrates Banchero and Jaylin Williams. Williams in particular being above Chet I don't understand, and that includes when Williams was getting some ASG buzz earlier. Williams is a very good player, but Chet is younger, pretty much just as good offensively (and more versatile) and a significantly better defender, and will likely be one of the 3-4 best defensive players in the NBA in his prime if he can add even a little more weight.
SGA is excellent, I don't really want to say anything bad about him...but I find it hard to rate someone that hasn't done anything in the playoffs before above someone like Tatum, who has dominated playoff series and done things like absolutely batter Kevin Durant on both ends of the floor, led his team past Giannis when Giannis was playing out of his mind, drop 51 points in a closeout game against Philly, etc.