Difference Making Games of 2018

If I have more time today than just being able to toss this post out there, I (or we....) should start a balancing thread since there will always be victories by the Sox that are due to other teams blown saves

Its a nice thought exercise, but without some comparison to what's "normal" or what other teams' blown-games record is, it will end up as just another post-hoc gripe repository.

santadevil said:

I've been thinking about this for a few days and it seems like at the end of every season, we look back at a few games that were lost, that should have been won. It's normally tough to remember those games and when they happened.

I hope we get to come back here at the end of September and say it didn't matter to the season. Maybe this thread sucks because it happens to each team, so no big deal. Opposite thread could be the "Games the Sox didn't deserve to win", which seems better anyway.

Click to expand...

Instead of "Blown Games," why don't you do a thread when a game has a certain win probability % - say 95% - and the other team wins? Not that it shows anything but it would be more balanced.

RS had a 97% chance of winning Game 1 but didn't.

wade boggs chicken dinner said:

Instead of "Blown Games," why don't you do a thread when a game has a certain win probability % - say 95% - and the other team wins? Not that it shows anything but it would be more balanced.

RS had a 97% chance of winning Game 1 but didn't.

Click to expand...

Its also an objective and comparable measure.

I would rather put these type games in the rear view mirror by the next day and move on, which means I will probably skip returning to this thread after this visit.

joe dokes said:

Its a nice thought exercise, but without some comparison to what's "normal" or what other teams' blown-games record is, it will end up as just another post-hoc gripe repository.

Click to expand...

Good point. Although I think we’d both agree that blowing a 4 run lead by giving up 6 in the 8th is definitely not normal.

The Yankees blew two games against Toronto and the Sox pulled out a few close wins so it’s a wash during the season IMO

While this does not exactly answer the question about blown saves, I think it adds dome perspective. It's based on a study I did several years ago using Retrosheet data (2000-2011) to look at when team had scored enough runs to win a game and is broken down by the percent of games in which a team had scored enough runs to surpass its opponents final score, by inning and by home and visiting team (I stands for Inning, X stands for all extra-inning games).

I Home--Vis
1 9.72--8.37
2 8.82--8.33
3 11.38--10.62
4 10.97--11.22
5 11.30---11.69
6 11.47--11.97
7 10.19--10.84
8 9.85--10.16
9 7.99--8.22
X 8.33--8.54

So, in the absence of a more complete study, one might investigate blown saves of teams losing more than about 8% of their games in the 9th inning. Other than that, it is interesting to see that home teams do most of their winning early: 29.92% in innings 1-3 as opposed to the visiting teams winning only 27.32% during those innings. This is home vs. visitor and home teams won 52.22% of those games.

Soxfan in Fla said:

Good point. Although I think we’d both agree that blowing a 4 run lead by giving up 6 in the 8th is definitely not normal.

Click to expand...

Normal was the wrong word to describe "comparing with how often it happens generally."

joe dokes said:

Its a nice thought exercise, but without some comparison to what's "normal" or what other teams' blown-games record is, it will end up as just another post-hoc gripe repository.

Click to expand...

Agreed. This thread is only useful if it includes every team in baseball. What’s the point of only looking at Red Sox fails without any context?

Jerry’s Curl said:

The Yankees blew two games against Toronto and the Sox pulled out a few close wins so it’s a wash during the season IMO

Click to expand...

Yes, we will lose 6-8 games we shouldn’t and win 6-8 we shouldn’t. It’s the beauty of a 162 game marathon. First one happening on opening day magnified it by 100.

geoduck no quahog said:

Agreed. This thread is only useful if it includes every team in baseball. What’s the point of only looking at Red Sox fails without any context?

Click to expand...

I think there is at least value in seeing which side of the ledger the Sox end up on so long as you look at it objectively. I like the idea of charting the 95% + win probability situations. Maybe also chart the 90% + ones as well as another data point.

geoduck no quahog said:

Agreed. This thread is only useful if it includes every team in baseball. What’s the point of only looking at Red Sox fails without any context?

Click to expand...

Actually, I'm not sure how useful it would be even if we included every team in the league.

However, it should provide some interesting anecdotes.

wade boggs chicken dinner said:

Actually, I'm not sure how useful it would be even if we included every team in the league.
However, it should provide some interesting anecdotes.

Click to expand...

I'm not sure it would be "useful," at least not in the predictive sense, in which it probably resembles extra-inning wins as interesting things not consistent from year to year. It could serve as a data point in a bullpen quality analysis. "Team X has blown twice as many 90% wins as any other team." I would assume it correlates with overall winning, but who knows, maybe its totally independent.

I think given last years frequent late heroics, Speier ran a couple of "sox have a .195* winning pct when down by 2 in the 8th *or later; league average is .120*" - type graphics during games.

(*all numbers invented).

To break it down into objective stats I think it could be instructive to look at win expectancy (WE%) after the 5th, 6th, 7th and 8th innings vs actual outcomes. In the opener that went 86.7% / 89.8% / 96.6% / 7.3% / loss. Game-by-game it's gone:

86.7% / 89.8% / 96.6% / 7.3% / loss
50.0% / 50.0% / 74.9% / 83.9% / win
78.4% / 82.0% / 86.7% / 83.9% / win
50.0% / 69.6% / 74.9% / 83.9% / win
92.3% / 94.5% / 98.4% / 98.5% / win
50.0% / 50.0% / 50.0% / 50.0% / win

with the average so far being:

67.9% / 72.7% / 80.3% / 67.9% / .833 win pct.

Using the End of the 7th vs Win Pct. comparisons, we should be able to gain a general measure of bullpen effectiveness. I'll try to add and keep up-to-date the similar stats of the competitor teams (NYY, TOR, CLE, HOU) if people are interested.

While useful and interesting overlap, they are not one in the same. It may not be the former, but it could be the latter, nonetheless.

I like the idea, UK, but what about stepping it up another notch and doing the following?

Every game begins with both teams' win expectancy (WE) = 0.5 and ends with one team's WE = 0.0 and the other WE = 1.0.

For losses:
If a team at any point in the game (no need to pick specific end-of-innings) reaches a WE of 0.9999999 and loses, that's an epic blown game. So the "Blow Factor" should be = 1.0. If they start off losing from the fist batter and never achieve a WE > 0.5, then they sucked but they didn't throw away anything resembling a win so the Blow Factor = 0.0. So one can look up the team's highest WE (hWE) of the game and calculate BF = (hWE - 0.5) x 2.0, which will run from 0.0 : 0.999999...

For wins:
Same concept but calculate lowest WE (lWE) and the "Comeback Factor" is CF = 1.0 - 2.0 x lWE, which also runs from 0.0 : 0.99999999....

If anyone has skills at scraping data from the web and could write code to collect game-by-game WE data and wants to collaborate, I'd be happy to do the analysis and make cool plots of BF and CF distributions for each MLB team. Again, it probably wouldn't be useful, but it could be interesting. Even just seeing what the shapes of the distributions look like could be neat.

Soxfan in Fla said:

Good point. Although I think we’d both agree that blowing a 4 run lead by giving up 6 in the 8th is definitely not normal.

Click to expand...

Yes and no. It's not something that's going to be all that common, but it's something that will likely happen to every team at least a couple times.

Rasputin said:

Yes and no. It's not something that's going to be all that common, but it's something that will likely happen to every team at least a couple times.

Click to expand...

It didn't happen here last year. Sox blew zero leads of three or more runs last season

Lose Remerswaal said:

It didn't happen here last year. Sox blew zero leads of three or more runs last season

Click to expand...

And here's a game it happened to the Astros that took me almost three whole minutes to find.
https://www.baseball-reference.com/boxes/KCA/KCA201706060.shtml

Good teams lose games in horrific ways every single year. Just like every year the world series winner has a 30 game stretch where they play mediocre or worse baseball.

If you're gonna track games the Sox lose in horrific fashion, you should track the ones the Sox steal as well.

dbn said:

I like the idea, UK, but what about stepping it up another notch and doing the following?

Every game begins with both teams' win expectancy (WE) = 0.5 and ends with one team's WE = 0.0 and the other WE = 1.0.

For losses:
If a team at any point in the game (no need to pick specific end-of-innings) reaches a WE of 0.9999999 and loses, that's an epic blown game. So the "Blow Factor" should be = 1.0. If they start off losing from the fist batter and never achieve a WE > 0.5, then they sucked but they didn't throw away anything resembling a win so the Blow Factor = 0.0. So one can look up the team's highest WE (hWE) of the game and calculate BF = (hWE - 0.5) x 2.0, which will run from 0.0 : 0.999999...

For wins:
Same concept but calculate lowest WE (lWE) and the "Comeback Factor" is CF = 1.0 - 2.0 x lWE, which also runs from 0.0 : 0.99999999....

If anyone has skills at scraping data from the web and could write code to collect game-by-game WE data and wants to collaborate, I'd be happy to do the analysis and make cool plots of BF and CF distributions for each MLB team. Again, it probably wouldn't be useful, but it could be interesting. Even just seeing what the shapes of the distributions look like could be neat.

Click to expand...

I've been collecting the peak and trough WE data for the losing and winning teams as well, but I'm not certain it tells us anything all that interesting. In order to have more epic fails you have to dominate earlier more often and vice versa. If anything I think it would show volatility or high runs scored/allowed.

By choosing distinct points in the game (i.e. WE before each of the last 4 innings) you avoid cherry-picking and can see how teams achieve their wins (i.e. by dominating early, coming back late, holding leads, etc.)

charlieoscar said:

While this does not exactly answer the question about blown saves, I think it adds dome perspective. It's based on a study I did several years ago using Retrosheet data (2000-2011) to look at when team had scored enough runs to win a game and is broken down by the percent of games in which a team had scored enough runs to surpass its opponents final score, by inning and by home and visiting team (I stands for Inning, X stands for all extra-inning games).

I Home--Vis
1 9.72--8.37
2 8.82--8.33
3 11.38--10.62
4 10.97--11.22
5 11.30---11.69
6 11.47--11.97
7 10.19--10.84
8 9.85--10.16
9 7.99--8.22
X 8.33--8.54

So, in the absence of a more complete study, one might investigate blown saves of teams losing more than about 8% of their games in the 9th inning. Other than that, it is interesting to see that home teams do most of their winning early: 29.92% in innings 1-3 as opposed to the visiting teams winning only 27.32% during those innings. This is home vs. visitor and home teams won 52.22% of those games.

Click to expand...

I like the concept here, but I don't understand your conclusions at all. 29% is less than a third of wins pre-won by the home team before the end of the 3rd inning. That's not most of the home team's wins.... Innings 3-7 have higher rates than the beginning or end innings. I'd look at that and say most games are won and lost in the middle innings.

Also it makes some sense that there would be a dip in the 9th inning rate specifically for home teams. There's 0 chance of pre-winning the game by scoring so many runs in that inning that the other team's later scoring is irrelevant.

Rasputin said:

And here's a game it happened to the Astros that took me almost three whole minutes to find.
https://www.baseball-reference.com/boxes/KCA/KCA201706060.shtml

Good teams lose games in horrific ways every single year. Just like every year the world series winner has a 30 game stretch where they play mediocre or worse baseball.

Click to expand...

Not an epic comeback, but one of all-time amazing regular season games by the Sox that I'm sure most here remember was this one, as part of a 2 -1 series win over the eventual world champs. The first inning was completely bonkers. Anyway, despite some rumblings by the Marlins coaches about sportsmanship (if I remember correctly), it had no meaningful bearing on their season (other than possibly galvanizing them).

dwainw said:

Not an epic comeback, but one of all-time amazing regular season games by the Sox that I'm sure most here remember was this one, as part of a 2 -1 series win over the eventual world champs. The first inning was completely bonkers. Anyway, despite some rumblings by the Marlins coaches about sportsmanship (if I remember correctly), it had no meaningful bearing on their season (other than possibly galvanizing them).

Click to expand...

Wow. I love the fact that Damon came within a home run of the cycle--in the first inning. I'm going to guess he was swinging for the fences the rest of the day (he had two more hits, a swinging strikeout, and a flyball to CF, so, yeah).

dwainw said:

Anyway, despite some rumblings by the Marlins coaches about sportsmanship (if I remember correctly)

Click to expand...

Sidebar:

The Marlins did complain - and were rewarded with an apology from the Red Sox's then-manager. "There's a right and wrong way to play the game and I feel we didn't play the game the right way totally." He also spoke to some members of his team about "mistakes" in judgment they made during the game. What kinds of mistakes? "They're not clearly defined. A lot of them are unwritten things, (but) we probably crossed the line in a couple of different areas."

Being an intelligent man, John Henry disagreed: "If we're supposed to stop trying to score, we should just put up a disclaimer on the scoreboard: 'You should go home now, we're not trying anymore.' You've got people giving up their Friday night to be at the ballpark, spending a couple of hundred bucks, and we're supposed to stop playing? The idea is to score runs. If not, then why are we out there?"

The next day on NESN, Jerry Remy said Boston's manager felt worse after the 25-8 win than he did after blowing three recent leads to the Phillies. And third base coach Mike Cubbage admitted he gave Florida extra outs. "It's a fine line. Guys want to score for the hitter. I kind of regret sending Walker. But it was a chance for outs, and they were having a hard time getting outs." (Boston's 14-run first inning ended with an easy out at the plate.)

Tony Massarotti and Steve Buckley of the Herald also decided the Red Sox had disgraced themselves with their disgusting run-scoring. Massarotti spoke of "a night for shame ... Red Sox ruthlessness on the basepaths ... no compassion ... tactless." Buckley added: "... your beloved Red Sox didn't exactly cover themselves in glory last night. They had this game in the bag in the first inning, and then spent the rest of the night taking the extra base, trying to score on close sacrifice flies ... boorish ... the Sox made ... asses of themselves."

Savin Hillbilly said:

Wow. I love the fact that Damon came within a home run of the cycle--in the first inning. I'm going to guess he was swinging for the fences the rest of the day (he had two more hits, a swinging strikeout, and a flyball to CF, so, yeah).

Click to expand...

That was a crazy good team. And while Todd Walker was probably better than given credit for, what the fuck was he doing batting 2nd while Billy Mueller was 8th. Should have been reversed.

I was there that night... as well as the next night. You know what made the apology so fucking stupid? The Sox blew a fucking 9-2 lead they had heading into the top of the 8th inning. You can never score enough runs in Fenway, especially not in that 2003 era. I fucking hate Grady Little.

There would've been nothing better than having Bill Belichick address that 14-run 1st.

Bump. Stole one back today

So, um, yeah.

AB in DC said:

So, um, yeah.

Click to expand...

Translation: we just scored six in the 8th to steal one from the Rays.

Rasputin said:

Translation: we just scored six in the 8th to steal one from the Rays.

Click to expand...

This had a real "all hands are ready" feel too. It was a professional come back.

Reverend said:

This had a real "all hands are ready" feel too. It was a professional come back.

Click to expand...

Yes. It was the kind of thing that you look back on after a successful season and say "see, the signs were there early on". Down 5 with two outs in the eighth inning of a game where not a whole lot went right (first poor outing by a starter all year, bullpen that couldn't keep runners off base, poor defensive play, injury to the hottest player on the team) and there was no fold up and quit in them.

Hard not to be excited about this team.

Red(s)HawksFan said:

Yes. It was the kind of thing that you look back on after a successful season and say "see, the signs were there early on". Down 5 with two outs in the eighth inning of a game where not a whole lot went right (first poor outing by a starter all year, bullpen that couldn't keep runners off base, poor defensive play, injury to the hottest player on the team) and there was no fold up and quit in them.

Hard not to be excited about this team.

Click to expand...

Wow. I feel even better about it now.

By the way, the Yankees had the bases loaded with nobody out, down one run in the bottom of the 12th. One Judge comeback GIDP and a Stanton K later, the game was over. These things happen.

Yankees’ win expectancy was 75% before the Judge gidp; Rays’ win expectancy was 98.5% before the Moreland double.

Red(s)HawksFan said:

Yes. It was the kind of thing that you look back on after a successful season and say "see, the signs were there early on". Down 5 with two outs in the eighth inning of a game where not a whole lot went right (first poor outing by a starter all year, bullpen that couldn't keep runners off base, poor defensive play, injury to the hottest player on the team) and there was no fold up and quit in them.

Hard not to be excited about this team.

Click to expand...

I had that thought about this game as well.

santadevil said:

Huge comeback yesterday. I'd say that puts the Sox 1-1 for the year.

Click to expand...

I think we need to count Thursday's win too. Down 2-0 heading into the bottom of the 9th. That's a pretty low probability.

Edit: per Fangraphs the Sox were down to 8.3% in that game. So it wouldn't meet the 95% threshold, but would meet the 90% threshold

For the first time this year the Sox have passed NY in Fangraph's expected wins (95.1 - 94.5), odds to win the AL East (49.3% - 44.1%) and to make the playoffs (90.4% - 89.1%). The Yankees still lead in odds to reach the ALCS (37.3% - 35.4%), win the Pennant (18.9% - 16.9%) and win a championship (11.7% - 9.4%). For the latter three categories Houston and Cleveland are currently 58.0% and 42.1% for making the ALCS, 34.4% and 20.1% for winning the Pennant and 22.7% and 11.4% for winning the WS.

Here's the expected wins and odds and they have changed since the pre-season for the top 7 AL teams:

Some interesting observations of the above:

• Angels have increased expected wins the most (+3.1) [the Ohtani effect?] and Sox next most (+2.6)
• Yankees have actually increased expected wins by +0.1 games despite opening 5-5. Sox have dropped their ROS win pct. by -.002 (probably because they have 9 fewer games to play against Florida clubs)
• Cleveland's 4-5 start has caused them to lose -2.8 expected wins, making them the only AL team expected to win > 81 games this year that has dropped since Opening Day.
• Sox have increased their odds of winning the division by the most of any AL team (+10.3% to 49.3%); the Angels have increased their odds of making the playoffs the most (+12.2% to 39.3%)
• The odds that one of the other 8 AL teams make the ALCS has dropped from 5.6% before Opening Day to 3.0% now.
• The odds that one of the other 8 AL teams wins the Pennant has dropped from 1.5% before Opening Day to 1.0% now.

This might be a fun way to look at how (un)expected each game win/loss turned out to be.

Below I plot the lowest (highest) Boston win expectancy in every win (loss), by game. So WE values above 0.5 are victories and the ordinate is the lowest the Red Sox's WE got before they came back, and values below 0.5 are losses and the ordinate is the highest the Red Sox's WE got before they blew it.



Nota bene that it is not expected that such data should be "normally" distributed, but on the right side of the plot I nonetheless show the Gaussian distribution "sigma" levels corresponding to the given percentage, since lots of us like to think in terms of sigmas. In fact, I think it's more useful to use sigma as the y-axis, so here is the same plot scaled that way, with the WE on the right now:



Also remember that this is a way to look at "comeback-yness" and "blowing it-yness", not how close a given game turned out to be. For example, game #3 was only a 3-2 victory, but sits in the middle of the y-axis because the Red Sox scored first and never trailed, so the lowest their WE got was 45.2 when the score was 0-0 heading into the bottom of the first; whereas game #8's 10-3 blowout sits higher because the Red Sox were down 0-2 at one point in the game and their WE was 31.0 at that point.

Not much to look at now, but if I keep up with this it might be fun to bin up the data, apply appropriate weightings and look at the distribution. Or not.

Soxfan in Fla said:

blowing a 4 run lead by giving up 6 in the 8th is definitely not normal

Click to expand...

Well, hopefully winning a game where we were down by 5 by scoring 6 in the 8th counterbalances that.

nevermind

uk_sox_fan said:

For the first time this year the Sox have passed NY in Fangraph's expected wins (95.1 - 94.5), odds to win the AL East (49.3% - 44.1%) and to make the playoffs (90.4% - 89.1%). The Yankees still lead in odds to reach the ALCS (37.3% - 35.4%), win the Pennant (18.9% - 16.9%) and win a championship (11.7% - 9.4%). For the latter three categories Houston and Cleveland are currently 58.0% and 42.1% for making the ALCS, 34.4% and 20.1% for winning the Pennant and 22.7% and 11.4% for winning the WS.

Here's the expected wins and odds and they have changed since the pre-season for the top 7 AL teams:

Code:

08-Apr   W    L      W%     EXPW     chg     EXPL    rosW%      chg       POFF      chg      DIV      chg     DOFF      chg      ALDS     chg      ALCS      chg      WS       chg
Astros   8    2    0.800    101.1   +0.9     60.9    0.613    -0.006     98.7%     -0.1%    95.2%    -0.7%    97.1%    -0.3%    58.0%    -1.0%    34.4%     -1.3%    22.7%    -1.3%
Tribe    4    5    0.444    93.6    -2.8     68.4    0.586    -0.009     91.1%     -5.5%    86.4%    -6.3%    88.9%    -5.9%    42.1%    -5.8%    20.1%     -3.1%    11.4%    -2.2%
BoSox    8    1    0.889    95.1    +2.6     66.9    0.569    -0.002     90.4%     +6.2%    49.3%   +10.3%    71.7%    +7.3%    35.4%    +4.9%    16.9%     +2.5%     9.4%    +1.4%
Yanks    5    5    0.500    94.5    +0.1     67.5    0.589    +0.006     89.1%     -0.6%    44.1%    -9.7%    69.9%    -3.9%    37.3%    -0.4%    18.9%     +0.4%    11.7%    +0.8%
Jays     6    4    0.600    86.0    +1.7     76.0    0.526    +0.006     42.7%     +5.6%     6.5%    +0.1%    23.3%    +2.6%     9.3%    +0.9%     3.7%     +0.6%     1.7%    +0.4%
Halos    7    3    0.700    85.6    +3.1     76.4    0.517    +0.008     39.3%    +12.2%     3.8%    +1.3%    20.1%    +5.8%     7.8%    +2.5%     2.9%     +1.0%     1.2%    +0.4%
Twins    4    3    0.571    83.3    +0.4     78.7    0.512      ---      28.9%     +0.2%    13.3%    +6.3%    19.9%    +3.1%     7.1%    +1.5%     2.1%     +0.4%     0.9%    +0.3%

Click to expand...

538 now has the Sox at 76 playoffs / 48 division / 9 WS to 69 / 36 / 8 for the MFY. Interesting to compare

uk_sox_fan said:

For the first time this year the Sox have passed NY in Fangraph's expected wins (95.1 - 94.5), odds to win the AL East (49.3% - 44.1%) and to make the playoffs (90.4% - 89.1%). The Yankees still lead in odds to reach the ALCS (37.3% - 35.4%), win the Pennant (18.9% - 16.9%) and win a championship (11.7% - 9.4%). For the latter three categories Houston and Cleveland are currently 58.0% and 42.1% for making the ALCS, 34.4% and 20.1% for winning the Pennant and 22.7% and 11.4% for winning the WS.

Click to expand...

I think it's interesting that the Yanks are judged more likely to get to the ALCS even though we're judged more likely to win the division. That seems odd.

Though, I seem to recall someone doing a study many years ago that suggested--among other things--that ideal postseason teams hit a lot of home runs. The idea being that a home run takes one bad pitch and sequential offenses take more than one bad pitch and bad pitches are rarer in the postseason.

Also, just in the realm of difference-making games, we're about to play three of them. Regardless of who gets out ahead or comes back or what, head to head games with the Yankees are going to mean a lot.

Maybe the games will be meaningful, but last year the Sox went 8-11 vs the MFY and still won the division by two games, so who knows. 8-11 likely won’t get it done this year though, I admit

Not sure where to put this... and not into starting a thread for a simple question: Is Devers considered a rookie this season? Would he be eligible for RoTY (obviously way too early to start considering that)? He's been off to a pretty nice start to the season so far and from what I've seen, his defense has been pretty solid at 3rd. I'm super bullish on him as a middle of the order bat for hopefully the next 10-12 years