Is 1 month too early to look at run differential?

[SoCalBrewfan]

There's several layers of things going on, too...the adjusted standings at BP suggest that the Brewers have played a bit better than their RS/RA would suggest, but also a bit worse than their record would suggest. Their equivalent runs scored and allowed are both a bit better than actual runs scored and allowd, so they've scored less often than raw hitting stats would suggest, and allowed more runs than raw opponent numbers would suggest. But we're still in the part of the season where it's difficult to separate fluctuations from trends. As N goes to infinity, 'things should even out,' but we're far from that point.

 

While the Brewers have had some bad innings by the back of the bullpen, they've also had good health for the most part, so it's dangerous to assume that we will play the rest of the season without significant innings from bad pitchers. Knock on Kerry Wood here, but last year should illustrate how easily the rotation can break down and overexpose lesser arms. And it's inevitable that the guys who have played well will hit some rough patches, too...I'm guessing that Cordero will allow a run sometime before June 1. At the same time, there are positions where things could still improve, and the Brewers do have some ready-made solutions if the guys starting now don't work out...in particular Braun should improve the third base situation even if Koskie doesn't return, and there are arms in AAA with promise.

 

Also at BP, three different simulations of the remainder of the season have the Brewers as most-probable winners in the Central. Their good performances against the division help quite a bit as well...several people here pointed to those division games as important in gauging the team's prospects. But of course there's a lot of baseball left to be played.

[dvoiss]

While R-RA is a fairly accurate formula for predicting wins that doesn't mean that the Brewers are going to start losing more it could simply mean that there are fewer 10-1 blowouts in their future and the differential could grow to more accurately reflect the team's record rather than the record changing to reflect the differential. In other words it's too early to tell.

[TooLiveBrew]

Quote:

---

You can predict a team's record over a season with only rs and ra pretty accurately.

---

Yes, and you'll come up with teams that "should" have made the playoffs or whatever, but the reality is different.

 

Russ, the reason I call run differential simplistic is because I believe that's just what it is. Used in concert with other factors, it can certainly be used to derive much more complicated stuff.

 

But to me, the bottom line is still embodied in the example of Saturday's game. Even though we were outscored by 10 x, we still only get one loss, and the 'Stros still only get one win.

 

I understand, too, how run diff. adds up over the course of a season and can come close to predicting a team's record, but the bottom line is that it's not how many games you could/would/should have won, it's how many you actually do win. Sorry if that comes off as condescending!

 

http://forum.brewerfan.net/images/smilies/smile.gif

 

 

p.s. -- Glad to see you back on these discussions rluz

[dadofandrew]

A couple games makes a huge difference after a month, yet nearly none after 162. It's a valuable stat, but like many others, means little after 2 dozen contests.

[rluzinski]

but the bottom line is that it's not how many games you could/would/should have won, it's how many you actually do win.

 

And who is saying otherwise? I've made two general points in this thread:

 

1. Given a large enough sample, run differential can tell you a lot about how a team might be expected to perform going forward. Since rs and ra correlated with wins pretty good (despite your previous claim), why shouldn't we assume some level of predictive ability?

 

2. 24 games is not a large enough sample for run differential to really matter.

 

That's it. If a team wins 120 games and the World Series, why would anyone care about run differential?

 

p.s. -- Glad to see you back on these discussions rluz

 

I took like a 5 hour break. That's all the will power I have apparently. http://forum.brewerfan.net/images/smilies/wink.gif

[rluzinski]

Len Kasper explains it well:

 

Quote:

---

The Cubs' Pythagorean record is 12-9, which means that what it roughly SHOULD be based on their run-scoring and run-prevention to this point. What does all this mean? Not a whole lot because it's early, but the bottom line is, beneath the record, the Cubs haven't been as bad as their record indicates. If they were getting blown out every day, that's a different story. In the big picture, if you outscore the opposition over a 162-game schedule, you should be over .500. Pretty simple. Now, the timing of when you score runs and give them up is the big key, but the hope here is that things will tend to even out a bit and the Cubs will eventually climb to .500. That is, assuming they maintain their current run differential pace.

---

 

LINK

 

You think Bill Schroeder knows the Brewers' pythagorean record? http://forum.brewerfan.net/images/smilies/smile.gif

[TuesdaysWithRillo]

Or maybe the Cubs are a bad team because bad teams blow close games cause of their bullpen, defense, etc.

[rluzinski]

If the Cubs finish the season 0fer in 1 run games, I'll give you a ZILLION dollars. If you are willing to admit the near impossibility of that, you will begin to appreciate what's being talked about here. Perhaps, because of the Cubs bullpen makeup or other factors, they aren't going to win as many close games as an average team, but there's no chance that they can stay winless in 1 run games for the whole season.

[endaround]

To underscore what Russ says. Assume the Cubs are really, really bad for whatever reason and only have a 30% chance of winning a one run game (keeping in mind a team winning 30% of their games overall is historically bad). Given 6 one run games in April lets project 30 more over the course of the season (a few on either side won't change much). The probability to go 0-30 then would be (.7)^30 or 0.00002.

[And That]

The probability to go 0-30 then would be (.7)^30 or 0.00002.

 

That's pretty good odds for a zillion dollars. Russ, can I get in on this action too?

[Team Canada]

To quote Kevin from The Office, if someone offers you 10,000 to 1 odds, you take it every single time.

[rluzinski]

That's pretty good odds for a zillion dollars. Russ, can I get in on this action too?

 

Provided the winner gets paid in "hyperbole bucks", sure.

[TuesdaysWithRillo]

If the Cubs finish the season 0fer in 1 run games, I'll give you a ZILLION dollars. If you are willing to admit the near impossibility of that, you will begin to appreciate what's being talked about here. Perhaps, because of the Cubs bullpen makeup or other factors, they aren't going to win as many close games as an average team, but there's no chance that they can stay winless in 1 run games for the whole season.

 

When did I ever say they would go 0fer in those games. I don't think they will. I just think they will have a below average record in one run games.

[sheethead]

Ummmm, I think the run differential is solved. Shouldn't be an issue for the rest of the year http://forum.brewerfan.net/images/smilies/smile.gif

[rluzinski]

When did I ever say they would go 0fer in those games. I don't think they will.

 

Well, then if we know the Cubs will win at least one, 1 run game this season, assuming their run differential doesn't get worse, we've predicted that the Cubs will improve their winning % to some degree. Great work everyone, that's a wrap!

[Lorax1]

Do people feel a little better about run differential after games like tonight?

[brewcrew44]

Heh, after one game like tonight our run differential is now very solid. I geuss that answers the question, "Is 1 month too early to look at run differential?" Yes. Yes it is.

[Athens Badger]

Exactly, I just came to post the same thing. +13 looks pretty decent at this point.

[Patrick425]

Haudricourt must be a brewerfan.net reader...

 

www.jsonline.com/story/index.aspx?id=599953

[DuWayne Steurer]

+13 does look a bit better, yes.

 

I don't think one should simply dismiss that the RD is low because of "blowout losses" though. The fact is, the Crew has been blown out a healthy number of times already, and if nothing else, it's an indicator that they've been lucky so far in the distribution of when those runs have been given up. That stuff will average out over the course of a season.

 

As long as Turnbow and Cordero can continue to pitch at an all star level, they'll continue to have a good record in close games, but if even one of those guys falters, that good luck could swing the other way pretty quick. It'd be nice to have some other quality options to pitch in a one run game in the 8th inning just so Turnbow doesn't have 75+ appearances over the course of the season.

[zzzmanwitz]

NL Run Differential

 

NYM +48

CHC +29

LAD +21

MIL +17

ATL +15

SD +12

CIN +9

ARI +7

PHI +4

FLA +2

SF +1

HOU -14

COL -22

PIT -28

STL -43

WAS -58

 

How a series can change things http://forum.brewerfan.net/images/smilies/smile.gif

[UeckerAddict]

+ 22 the last four days. How about we span that stat out for the rest of the season and see what we come up with?

[Team Canada]

You gotta believe both PIT and CHC are in for a reversal of fortunes. That's just too big to last.

[soup can for prez]

Quote:

---

Heh, after one game like tonight our run differential is now very solid. I geuss that answers the question, "Is 1 month too early to look at run differential?" Yes. Yes it is.

---

 

I could not agree more....

[TuesdayWithRillo]

Well, then if we know the Cubs will win at least one, 1 run game this season, assuming their run differential doesn't get worse, we've predicted that the Cubs will improve their winning % to some degree. Great work everyone, that's a wrap!

 

What if the Cubs don't play any one run games from here on out? That is possible to although very improbable. Just pointing out flaws in your theory since we don't know how many one run games they will be in.