Schoenfield: O's Starting to Look Like Team of Destiny

[McNulty]

Hugely. I had a 20 minute conversation with a guy last night just detailing the season, the factors at play, what's likely going forward.

Its not a conversation when the other person never gets a chance to talk.

This article was terrible BTW.

[TiredofLosing20]

I really don't get people's animosity towards the Pyth or the people looking at it and being worried that this team is not as good as it's record. It's pretty simple really.. If you look the Angels. They are performing exactly to their Pyth (67-62). The O's are 9 games better than they Pyth. To take math out of it, which team would you suggest has a better chance of reproducing their season next year provided the same rosters? How confident are you that this roster kept largely intact without significant upgrade could go out and win 90 games next year (which is basically the pace they are on)? Would you have more confidence in the Angels as constructed today to get there?

Now obviously rosters change and that clearly highlights the issue with using the Pyth to predict the future. But to me the strength of the Pyth is that it gives some context to your history or your W-L, thats all.

[Mizerooskie]

I really don't get people's animosity towards the Pyth or the people looking at it and being worried that this team is not as good as it's record. It's pretty simple really.. If you look the Angels. They are performing exactly to their Pyth (67-62). The O's are 9 games better than they Pyth. To take math out of it, which team would you suggest has a better chance of reproducing their season next year provided the same rosters? How confident are you that this roster kept largely intact without significant upgrade could go out and win 90 games next year (which is basically the pace they are on)? Would you have more confidence in the Angels as constructed today to get there?

Now obviously rosters change and that clearly highlights the issue with using the Pyth to predict the future. But to me the strength of the Pyth is that it gives some context to your history or your W-L, thats all.

I actually think Pyth gives no context at all. It doesn't take roster changes into context, doesn't take unbalanced scheduling into context, among other things.

As a predictor, it leaves a lot to be desired.

[Moose Milligan]

Hugely. I had a 20 minute conversation with a guy last night just detailing the season, the factors at play, what's likely going forward.

Its not a conversation when the other person never gets a chance to talk.

But that guy was surely impressed by Lucky Jim's large vocabulary and wit.

[Lucky Jim]

But that guy was surely impressed by Lucky Jim's large vocabulary and wit.

WTF is wrong w/ you?

[Moose Milligan]

WTF is wrong w/ you?

Haha, nothing. Relax man, I'm just poking fun at you, you're one of my favorite posters.

[CA-ORIOLE]

I actually think Pyth gives no context at all. It doesn't take roster changes into context, doesn't take unbalanced scheduling into context, among other things..

Then it does give some context.

As a predictor, it leaves a lot to be desired

Has anybody disputed that?

[Fan4Life]

I actually think Pyth gives no context at all. It doesn't take roster changes into context, doesn't take unbalanced scheduling into context, among other things.

As a predictor, it leaves a lot to be desired.

Amen!!!

[SilentJames]

Man I hate to do this, but this is pretty much the point I was trying to make about our run differential last month - and got lambasted for it.

We were good enough to build the record through the first third, we were lucky to stay in it the second third, the last third is really the product of DD, Showalter and the true talent of this team returning.

[Pedro Cerrano]

I've said several times that using past run differential, largely caused by players who aren't on the team anymore, to predict future success, is dangerous.

[TiredofLosing20]

I actually think Pyth gives no context at all. It doesn't take roster changes into context, doesn't take unbalanced scheduling into context, among other things.

As a predictor, it leaves a lot to be desired.

Did you even read my post?

Of course it doesn't take roster change. I specifically pointed that out as an issue. I also stated there are general issues with using it to predict future. But there is clearly a difference between a team that is blowing teams out versus one that wins a historical number of 1 run games. But certainly, ignore it if it contains no value to you.

[CA-ORIOLE]

Lol at this tweet. People are fixated on the 1-run thing now.

I've contended all along that's what it's really about. Call it what you want, but if it's something that's not projectable, I'll group it in with "luck". From what research I've read there is no correlation between winning a high percentage of one run games and strong BP performance. If there was I'm sure Buck would have executed the one run strategy long before now.

[callahan09]

The problem with pythag as a predictor is that something like this can happen (well, it's debatable whether it ever actually COULD happen, but it "could" happen):

Through the entire 162 game season, a team's offense scores 1 run per game.

For the first third of the season, a team's pitching staff gives up 10 runs per game.

The team goes 0-54 to start the season, having allowed 540 runs, and scored only 54.

For the next two thirds of the season, they replace their pitching staff with 5 aces the likes of which have never been seen before, giving up 0 runs per game for the next 108 games.

They finish the season 108-0, allowing 0 runs and scoring 108. That's a total record of 108-54, 540 runs allowed and 162 scored.

If you just look at the pythag for this team in the end of things, you'd have to predict that they will only win 16 games next season, assuming nothing changes. Of course,given the eyeball test, you'd come to almost the exact opposite conclusion and have to predict that they'd be undefeated in the upcoming season.

Obviously this is a ridiculously extreme example, but the numbers work to prove that pythag doesn't tell enough of the story to be used a predictor.

[CA-ORIOLE]

I've said several times that using past run differential, largely caused by players who aren't on the team anymore, to predict future success, is dangerous.

Not surprising, it's been psoted on here many times over.

[Lucky Jim]

Man I hate to do this, but this is pretty much the point I was trying to make about our run differential last month - and got lambasted for it.

We were good enough to build the record through the first third, we were lucky to stay in it the second third, the last third is really the product of DD, Showalter and the true talent of this team returning.

Except the point was wrong. Because we had a pretty much equal contribution to our above-pythag performance from all parts of the season. Including August, where before last night we were 15-8 and should have been 12-11 and after last night are 16-8 and should be 13-11. That's three wins in 24 games. A ratio of a pythag "win above" per 8 games. Which would put us at 20 games above over a whole season.

You should hate to do this because your run differential post was mathematically and logically flawed.