Comments on: Regression in LA?

By: JoeyO

JoeyO — Sat, 10 Apr 2010 02:11:27 +0000

“JoeyO, same questions”

It was the CHONE projections, I cant change the results (no matter how much screaming Xeifrank did insisting that I somehow must).

By: B N

B N — Sat, 10 Apr 2010 01:34:04 +0000

Wins Above Replacement Manager? WARM?

By: Nick

Nick — Mon, 01 Feb 2010 21:48:08 +0000

its funny you say that when most would say the reason they outperform their projections is the very man whose name you can’t spell.

and izturis over wood = toast shows how little you really know about baseball.

By: Nick

Nick — Mon, 01 Feb 2010 21:28:51 +0000

this is for NEPP but i dont see a reply in his box.

“It makes you wonder if his offense is somewhat a creation of playing in the PCL (a notorious hitters league).”

typical fringe fan comment. take a look at that dudes numbers every single year hes played pro ball. that’s all i am going to say.

“He’s struggled in his brief cameos in the Majors so far.”

its hilarious that you draw conclusion from his big league playing time. that sample size problem is like stats 101.

By: Xeifrank

Xeifrank — Fri, 15 Jan 2010 07:04:28 +0000

No argument here. It appears you do not understand my point. Put the 20 players at 0.0 WAR, which is fine by me. Then you need to add a negative offset for the inevitable negative hitting WAR. Because you are guaranteed that in the end there will be negative hitting WAR. Guaranteed! More for NL teams due to the pitchers batting. Take a look at one of the teams for 2009 on Fangraphs. Click on the ‘value” tab and add up all the negative hitting WAR. It must be accounted for, even if you cannot predict from which player(s) it will come from. Otherwise, you end up with weird predictions that have every team in the AL West winning 84 to 90 games and the A’s winning 87. You will also most likely end up with a league wide projected winning percentage well above .500.
vr, Xei

By: Xeifrank

Xeifrank — Fri, 15 Jan 2010 05:31:17 +0000

hahah Jacob, I tried to avoid the flame war, but was followed here. What can I do? Good question about the balanced results. I calculated for all 30 clubs and am slightly over .500 when adding up all win/loss records. I can live with that margin of error. If I were doing it for a college thesis, of course I’d tighten the strings a little bit. It’s quite likely that if you project the worst team in the AL West to have 84 wins and have the A’s at 87 wins that you are likely not to be very balanced using the same method for all thirty teams. If Vegas set the over/under on A’s wins at 87, there would be a lopsided amount of money bet on the under, no? Thanks for the thoughtful question(s).
vr, Xei

By: DJacobs

DJacobs — Fri, 15 Jan 2010 04:51:48 +0000

Wow guys. I think you don’t realize that you’re in two different conversations.

Xei, are you calculating for all 30 clubs and getting a balanced result?

JoeyO, same questions.

By: JoeyO

JoeyO — Fri, 15 Jan 2010 02:51:55 +0000

I guess… Well, that is, it proves your point if you’re only point is that you dont want to believe the CHONE projections.

Of course that also doesn’t mean anything, and the CHONE projections still remain as stated.

By: JoeyO

JoeyO — Fri, 15 Jan 2010 02:38:00 +0000

You’re kidding right?

Ok fine, ignore the first part then. Here:

CHONE Projections off current 25-Man rosters:
Rangers Total WAR = 41.9, Rangers rough record of 90-72
Athletics Total WAR = 39.0, Athletics rough record of 87-75
Mariners Total WAR = 37.2, Mariners rough record of 85-77
Angeles Total WAR = 36.3, Angels rough record of 84-78

Any remaining playing time goes to Replacement Level Players. Replacement Level Players are assumed to have a Zero WAR, it’s the entire basis of RAR/WAR. Sure any season some might be higher, some might be lower; but they have to be assumed to be Zero.

Now I have a question for you. Say there are 20 such replacement players needed. What difference is there between these two examples of said 20 players?

Example A:
20 at 0.0

Example B:
2 at +0.5
2 at +0.3
2 at +0.2
3 at +0.1
2 at 0.0
3 at -0.1
2 at -0.2
2 at -0.3
2 at -0.5

Truth is there is no difference between those two, is there? But if your insistence is solely that those 20 replacement players must be inserted in that (or some similar) random manor which shows some players listed as a negative, feel free to knock yourself out. And in the end, the results will correctly stay

By: Xeifrank

Xeifrank — Fri, 15 Jan 2010 02:21:22 +0000

Rangers Total WAR = 41.9, Rangers rough record of 90-72
Athletics Total WAR = 39.0, Athletics rough record of 87-75
Mariners Total WAR = 37.2, Mariners rough record of 85-77
Angeles Total WAR = 36.3, Angels rough record of 84-78

87 wins for the A’s. This proves my point.
vr, Xei