OPS During Wins/Losses

[camisadelgolf]

It is to be expected that a player's numbers will be worse in games in which his team lost. Generally, the decline in percentage appears to be around 30-40%. However, I was looking at the career numbers for the Reds, and Jay Bruce's decrease in production during losses was such a huge difference that I felt the need to start a thread about it.

Name		WinOPS	LossOPS	OPS Lost %
Jay Bruce	1.019	0.566	80.04%
Laynce Nix	0.851	0.543	56.72%
Drew Stubbs	0.887	0.582	52.41%
Miguel Cairo	0.793	0.562	41.10%
Paul Janish	0.728	0.516	41.09%
Scott Rolen	1.001	0.718	39.42%
Joey Votto	1.109	0.803	38.11%
Jonny Gomes	0.925	0.675	37.04%
Orlando Cabrera	0.819	0.602	36.05%
Ryan Hanigan	0.846	0.633	33.65%
BrandonPhillips	0.853	0.641	33.07%
Corky Miller	0.646	0.487	32.65%
Ramon Hernandez	0.836	0.638	31.03%

During wins, only Votto is better than Bruce. During losses, nearly the entire team is better than Bruce. Can we draw any conclusions from this? I don't know. The sample size is somewhat small (only 1,412 PAs for Bruce), but I really think this could say something about Jay Bruce.

[Hoosier Red]

It is to be expected that a player's numbers will be worse in games in which his team lost. Generally, the decline in percentage appears to be around 30-40%. However, I was looking at the career numbers for the Reds, and Jay Bruce's decrease in production during losses was such a huge difference that I felt the need to start a thread about it.

Name		WinOPS	LossOPS	OPS Lost %
Jay Bruce	1.019	0.566	80.04%
Laynce Nix	0.851	0.543	56.72%
Drew Stubbs	0.887	0.582	52.41%
Miguel Cairo	0.793	0.562	41.10%
Paul Janish	0.728	0.516	41.09%
Scott Rolen	1.001	0.718	39.42%
Joey Votto	1.109	0.803	38.11%
Jonny Gomes	0.925	0.675	37.04%
Orlando Cabrera	0.819	0.602	36.05%
Ryan Hanigan	0.846	0.633	33.65%
BrandonPhillips	0.853	0.641	33.07%
Corky Miller	0.646	0.487	32.65%
Ramon Hernandez	0.836	0.638	31.03%

During wins, only Votto is better than Bruce. During losses, nearly the entire team is better than Bruce. Can we draw any conclusions from this? I don't know. The sample size is somewhat small (only 1,412 PAs for Bruce), but I really think this could say something about Jay Bruce.

I think the fact that Bruce and Stubbs are so high on the list has to do with the large gaps between peaks and valleys that both experienced. In games that occurred while Bruce and Stubbs were on their collective hot streaks the Reds were practically unbeatable. When both were in their post all-star break doldrums, the Reds were utterly beatable.

But I'm not sure the correlation equals causation in either direction.

[RedsManRick]

It is to be expected that a player's numbers will be worse in games in which his team lost. Generally, the decline in percentage appears to be around 30-40%. However, I was looking at the career numbers for the Reds, and Jay Bruce's decrease in production during losses was such a huge difference that I felt the need to start a thread about it.

During wins, only Votto is better than Bruce. During losses, nearly the entire team is better than Bruce. Can we draw any conclusions from this? I don't know. The sample size is somewhat small (only 1,412 PAs for Bruce), but I really think this could say something about Jay Bruce.

Be careful which way you draw the causal arrow. Those performances aren't happening as a result of the Reds losing games. Rather the Reds lost games when players performed at that level -- along with a number of other factors (pitching and defense).

We know that in a given game, no one players' performance counts for more than another's (scaled to PA) and this reflects that. As a team, the Reds lose when they struggle to score runs and they win when they score a lot of runs. The combination of players who struggle or succeed isn't really all that important compared simply to how many players struggled or succeed.

That Bruce's split is so extreme simply shows that his struggles and success were most strongly correlated with the struggles and success of his teammates. Not that his struggles and success were particularly influential. There's really nothing here which suggests that the team struggled because of Bruce, that Bruce struggled because of his teammates, or if there is any meaningful causal relationship at all.

Rather than looking for some causal relationship between Bruce and his teammates, I think the more likely explanation is a strong correlation of offensive performance and opposing pitcher quality. Facing a good pitcher, everybody tends to perform worse. Facing poor pitchers, everybody tends to perform better.

In this model, one possible explanation for Bruce's extreme split is that his great talent lets him destroy bad pitching but his still-in-development plate approach makes him particularly susceptible to good pitching. Additionally, given the way OPS is calculated, players whose OPS is disproportionately comprised of SLG will be more strongly impacted by this effect. So guys who tee off on bad pitching are likely to experience the biggest splits.

		OPSloss	OPSwin	Split	Var
Jay Bruce	.566   1.019	.453	-.196
Laynce Nix	.543	.851	.308	-.051
Joey Votto	.803   1.109	.306	-.049
Drew Stubbs	.582	.887	.305	-.048
Scott Rolen	.718   1.001	.283	-.026
Jonny Gomes	.675	.925	.250	.007
Miguel Cairo	.562	.793	.231	.026
Orlando Cabrera	.602	.819	.217	.040
Ryan Hanigan	.633	.846	.213	.044
Brandon Phillip	.641	.853	.212	.045
Paul Janish	.516	.728	.212	.045
Ramon Hernandez	.638	.836	.198	.059
Corky Miller	.487	.646	.159	.098
Average		.613	.870	.257	.000

Var is the difference between the average split and the player's split. Check it out. See a pattern? Big slugging guys at the top, small slugging guys at the bottom. Put simply, the bigger they are, the harder they fall. I think you'll find that SLG or ISO accounts for a good amount of what you're seeing here.

[camisadelgolf]

Be careful which way you draw the causal arrow. Those performances aren't happening as a result of the Reds losing games. Rather the Reds lost games when players performed at that level -- along with a number of other factors (pitching and defense).

We know that in a given game, no one players' performance counts for more than another's (scaled to PA) and this reflects that. As a team, the Reds lose when they struggle to score runs and they win when they score a lot of runs. The combination of players who struggle or succeed isn't really all that important compared simply to how many players struggled or succeed.

That Bruce's split is so extreme simply shows that his struggles and success were most strongly correlated with the struggles and success of his teammates. Not that his struggles and success were particularly influential. There's really nothing here which suggests that the team struggled because of Bruce, that Bruce struggled because of his teammates, or if there is any meaningful causal relationship at all.

Rather than looking for some causal relationship between Bruce and his teammates, I think the more likely explanation is a strong correlation of offensive performance and opposing pitcher quality. Facing a good pitcher, everybody tends to perform worse. Facing poor pitchers, everybody tends to perform better.

In this model, one possible explanation for Bruce's extreme split is that his great talent lets him destroy bad pitching but his still-in-development plate approach makes him particularly susceptible to good pitching. Additionally, given the way OPS is calculated, players whose OPS is disproportionately comprised of SLG will be more strongly impacted by this effect. So guys who tee off on bad pitching are likely to experience the biggest splits.

		OPSloss	OPSwin	Split	Var
Jay Bruce	.566   1.019	.453	-.196
Laynce Nix	.543	.851	.308	-.051
Joey Votto	.803   1.109	.306	-.049
Drew Stubbs	.582	.887	.305	-.048
Scott Rolen	.718   1.001	.283	-.026
Jonny Gomes	.675	.925	.250	.007
Miguel Cairo	.562	.793	.231	.026
Orlando Cabrera	.602	.819	.217	.040
Ryan Hanigan	.633	.846	.213	.044
Brandon Phillip	.641	.853	.212	.045
Paul Janish	.516	.728	.212	.045
Ramon Hernandez	.638	.836	.198	.059
Corky Miller	.487	.646	.159	.098
Average		.613	.870	.257	.000

Var is the difference between the average split and the player's split. Check it out. See a pattern? Big slugging guys at the top, small slugging guys at the bottom. Put simply, the bigger they are, the harder they fall. I think you'll find that SLG or ISO accounts for a good amount of what you're seeing here.

I agree with everything you said, but Jay Bruce's split is so extreme that I'm wondering if there's more to it. His ability to feast on bad pitching is probably the biggest factor, but could part of it be mental? Does he give up when his team is losing? Does he put too much pressure on himself? Does he see more pitches he has trouble with when his team is behind?

[RedsManRick]

I agree with everything you said, but Jay Bruce's split is so extreme that I'm wondering if there's more to it. His ability to feast on bad pitching is probably the biggest factor, but could part of it be mental? Does he give up when his team is losing? Does he put too much pressure on himself? Does he see more pitches he has trouble with when his team is behind?

Those are perhaps fair questions to ask, but they certainly won't be answered by the data. This is actually a great example of how stats and scouting can work well together. The stats can help you ask better questions while the answer requires a bit more nuance.

[camisadelgolf]

Those are perhaps fair questions to ask, but they certainly won't be answered by the data. This is actually a great example of how stats and scouting can work well together. The stats can help you ask better questions while the answer requires a bit more nuance.

Yes. So . . . does anyone here have any thoughts based on what you saw from Jay Bruce this year?