12/30/2010 10:13 AM
And, by the way, it's the same in HBD.    It kills me when someone posts "What's wrong with my player?  He's hit .300 for 5 seasons in a row and, despite no rating changes, he's only hitting .240 this year after 400 AB."
12/30/2010 11:10 AM
Posted by MikeT23 on 12/30/2010 10:11:00 AM (view original):
I will assume you are part of the nerd faction. 

I'll try to break it down for you in a language you understand.   .250 is considered mediocre, .300 is considered good.   The difference between the two is about 1 hit PER WEEK over the course of a season.   Blooper, broken bat dribbler, poor jump by a fielder, etc, etc. can turn a player from an also ran to an All-Star in a season.   An AB, game, week, season are all small sample sizes. 

You can work with career averages and get an idea of probabilities but you can't figure out the outcome.   Because the outcome is from that small sample size of an AB or game.

That's why the game of baseball can't be wrapped up in nice, tidy number by statnerds. 
I agree with the point that you can do all the math you want but outcomes will often deviate from expectations, but i do believe that having a strong understanding of probabilities based on sufficient data can improve the odds of making good decisions. For example, how many threads have we seen about "which SS should i play, the glove or the stick". The obvious response is the one that helps your team the most. In that calculation, the offensive impact isn't too difficult to determine, (roughly, of course), what's harder is to estimate the defensive contribution. If you can get a decent handle on how many runs one guy will save versus another in the field it will help in deciding which player brings more to the table.
12/30/2010 11:40 AM
With 41 seasons under your belt, I'm sure you "eyeball" a guy and get an idea.  There's just no way you can say 5 negative plays equates to 2.1 runs.   There are too many moving parts.   Really all you have to determine is if 5 negative plays and a .300 average is better than 5 positive plays and a .280 average.   And, of course, that .020 difference in average is 1 hit every 12-13 games.    And, of course, you know a +/- of 10 is 1 "hit" added/removed every 16 games.    That's very simplistic as all plays, good and bad, are not the same.   Which is why the game can't be completely defined by nice, tidy numbers.   
12/30/2010 12:26 PM
I am not saying that 5 negative plays will always, definitively, allow an extra 2.1 runs. I am saying that if a team allows 50 minus plays over the course of a season, those 50 missed plays will USUALLY be worth ABOUT 30-35 extra runs allowed. Sometimes it will be more, sometimes less, but for planning purposes, that will get you closer to what's likely to occur than any other method i've seen.

To your example, those guys would be close enough from a production standpoint that i would look at other factors to determine which i'd prefer - age, contract, dur, health, etc.

Just because we can't completely define the game with numbers, doesn't mean we can't gain an edge by trying to refine our understanding of the odds.
12/30/2010 1:20 PM
Just for the sake of discussion, I'll post this guy:

Hardball Dynasty – Fantasy Baseball Sim Games - Player Profile: Brian Mattingly

I see him, defensively, as a very good 1B or a below average LF.  For S9, I moved him to LF for the season.  I picked up a basher who was a legit 1B so the move had to be made.   We won 91 games in S8, 101 in S9.  My pitching, largely the same, went from 3.92 ERA to 4.48(league average dropped .28).  My positive/negative went from 88/21 to 66/42.  Fielding percentage remained constant at .986(52 unearned runs in S8 vs. 55 in S9).   So the "logical" explanation is that the 43 total change in positive/negative plays was the culprit.   Now, can I really say that's the reason for the extra 70someodd runs?  The stats would indicate "yes".  But I know that the move of Mattingly, which weakened two positions in the field, didn't cost me 70 runs(or .84 ER based on the league average change).    But I believe adding that bat helped me win 10 more games. 

So, would I be wise to tie myself up in stats that indicate I'm destroying my pitching staff by using a poor LF or should I accept my 10 extra wins as confirmation that I'm doing the right thing?
12/30/2010 8:01 PM
Sorry for the delay - holiday travel on 95 - nuthin' finer.

So first, context . . . in S8 your RS/ RA was 797/687, per Bill James pythag you should have won 92 or 93, you did win 91. Smack in the middle of the bell curve. In S9 the RS/RA was 900/774, proj wins 93, actual wins 101. This implies some luck may have been involved - the very deviance from "tidy numbers" that you wisely warn folks to be on the lookout for. Your 1 run games and extra innings advantages seem to support this. Your "closer" was, if anything, a bit less effective in S9 than S8, which again suggest you might have caught a break or two.

How did you get more offense? Primarily by adding two key players, Baek and Sanchez who probably accounted for about 2/3rds of the additional run production.
Why did you give up more runs? At a quick glance, your pitching staff, as you acknowledge seems to be mostly the same names, though a couple of them seem to have had unusually successful seasons (compared to career norms) in S8 (Moreno/Valdes).
Role of defense? Your error rate was virtually the same, no surprise that unearned runs were also about the same. However, your "net" plus plays dropped by 43, I would assume that accounts for approximately 30 runs of the additional 87 given up from S8 - S9.

Mattingly's role in this? Hard to say. In your post you say you "moved him to LF for the season(S9)", but it looks like he started 120 games in LF in S8 vs. 152 games in LF in S9. Based solely on production, I have him worth about 130 - 135 RS. Adjusted for defense, I'd say he saves 4 runs at 1B and "costs" 7 runs in LF. Less than one Win either way.

So, my 41 season, quant/number nerd analysis? Your additions of Baek and Sanchez over Gonzalez and Javier at 1B and 3B were enough to compensate for a regression to the mean of your pitching staff and you caught a couple breaks to get the extra few wins. As for Mattingly? With less than 1 Win expected difference, I'd put him in Left or 1B as your alternatives allowed.

Just my two cents. One of the great things about this game (as you note) is that know one we know has the secret formula to crack the black box, so the fun for me is in trying to tease out as much of it as I can.

12/31/2010 12:56 AM
"Really all you have to determine is if 5 negative plays and a .300 average is better than 5 positive plays and a .280 average."

Yes... yes if there was only some way that one could determine which one was better. That way would probably use numbers and stuff and some sort of counting machine. Let me find my abacus.

"Which is why the game can't be completely defined by nice, tidy numbers."

Well, for real-life baseball, you have a point. However, HBD is a computer game. Thus, it literally can and is defined by nice, tidy numbers.

I have an idea, let's all go to the casino and stand on 16 when the dealer shows 10. Because hey, it totally worked that one time.
12/31/2010 8:16 AM (edited)
Sorry, jtr, it can't unless you have some secret formula that provides you with all sorts of info that the rest of us don't have.  Of course, with exactly a .500 record in HBD, it doesn't appear that you're using this secret formula for HBD success.

dwool, I'm well aware that I scored more runs and that, in itself, will account for more wins.   My bigger question is why did my ERA, based on league average, rise almost one full run?   Valdes' first pro season was S8 so I don't think we can have a "career norm".   Moreno was actually better in S7 than he was in S8 and better in S8 than S9(without rating changes).   What you probably don't know is that I joined in S6 and concentrated on picking up defensive players because I couldn't get offensive players.  So my D has regressed, ratings-wise, in each season(S6-S9) as I acquired more offensively-minded players.  Etherton's ratings have remained the same but he too has had a performance decline from S7-S9(very similar to Moreno).    In fact, that's a pretty common theme with all the pitchers who have been with me all three of those seasons.   So, can you quantify the effects on "career norms" for pitchers with a declining defense?   I actually find this pretty interesting because I have 3 full seasons of data and I know exactly how I built the team(and the defensive stats show it).   

To give you a little more info, I took over a downtrodden team(never .500 in it's first 5 seasons) in an established world.   I've found that the "best" way to be competitive is to collect defensive castoffs(good D/poor bat) and try to win the 2-1, 3-2 games while adding a hitter here and there.   So, over the course of those 4 seasons, the team went from D to O at the "hitting" positions.    While my fielding percentage stayed the same, the +/- play differential declined.  And the pitchers who were there the entire time went from "overperformers" to "career norms".   What effects does the defense really have? 
12/31/2010 9:24 AM

Here are the pitchers I'm looking at:

Hardball Dynasty – Fantasy Baseball Sim Games - Player Profile: B.C. Mota
Hardball Dynasty – Fantasy Baseball Sim Games - Player Profile: Jorge Moreno
Hardball Dynasty – Fantasy Baseball Sim Games - Player Profile: Howie Etherton
Hardball Dynasty – Fantasy Baseball Sim Games - Player Profile: Wilfredo Sanchez
Hardball Dynasty – Fantasy Baseball Sim Games - Player Profile: Gabby Wilkerson

All threw a significant amount of innings(although still a SSS) in each S7, S8, S9.

My fielding percentage stayed constant.   The only difference is the change in +/- plays.    How significant are they?   If you look at the numbers of those five pitchers, pretty damn significant.

 

12/31/2010 9:49 AM

Totals for all 5:

 

Inn

H

R

ER

HR

BB

SO

WHIP

ERA

7

594

518

238

220

61

162

445

1.15

3.34

 

 

 

 

 

 

 

 

 

 

8

606

610

280

255

66

184

419

1.31

3.79

 

 

 

 

 

 

 

 

 

 

9

487

505

265

250

63

190

341

1.43

4.62

12/31/2010 9:59 AM
What was your DER those three seasons, Mike?
12/31/2010 2:13 PM
His DER was .721, .714, and .709 in S7, S8, and S9 respectively.

I think you were a bit unlucky defensively in S9. Without checking, I would bet that most teams with an above-average fielding percentage and a +24 +/- will not underperform their FIP by 0.4. Interestingly enough, although you had a +24 +/-, your DER in S9 (.709) was worse than league-average, which was .714 for that season. That could have a lot to do with a ground-ball heavy staff though, as ground balls have higher batting averages than fly balls that stay in the yard.

I do believe that outfield +/- is more important than infield. Your infield defense was quite good, you were just terrible in left and right field. Combine that with a home park that already boosts doubles and triples, and it's probably not a great competition. As you mentioned though, it was a worthwhile trade-off, because the offensive boost easily made up for it.

Another factor is catcher defense.

In S8, you threw out runners at a .373 clip. In S9, that number was reduced to .223. In 130 attempts, that's a difference of about 20 outs. For simplicity's sake, I'll ballpark each of these outs as worth about .4 runs, and say that your catchers were 8 runs worse at throwing out runners in S9. Additionally, your "weighted pitch calling" (the pitch rating of your catchers times their share of innings they caught) dropped from 78 to 71. You've mentioned before that you think 10 points in pitch calling = .1 in ERA, so call that .07 in ERA, and that's another 9ish runs. So overall, your catcher defense could be estimated to be in the ballpark of 17 runs worse in S19, which seems to put your non-catcher defense at about 50 runs worse from S8 to S9.

So, combine a weaker defense not optimally suited for the park, with catchers significantly worse at throwing out runners and calling games, and add in a little bit of bad luck and a heavy pitcher's park (Santa Cruz) in your division being swapped for a slight hitter's park (Salt Lake) and you have a recipe for a pretty big increase in ERA.
12/31/2010 3:04 PM
1/1/2011 10:21 AM
Thanks for the link boog, that was a mych more elaborate expression of what I was crudely trying to get at.
1/1/2011 3:00 PM
JTR, while much of what you say is correct, it has to be more than just a "bit of bad luck".    And swapping out one park in my division really doesn't cut it.  That's 5 more games and, on the whole, the league ERA went down .028. 

I'll accept that some of it could just be a "random result".    However, there's a gorilla in the room dressed as 5 pitchers, with virtually no rating changes over three seasons, that got worse as the +/- play differential got worse.   We can't say "That's not a gorilla" while he smacks us in the head.

of 7

Terms of Use Customer Support Privacy Statement

Popular on WhatIfSports site: Baseball Simulation | College Basketball Game | College Football Game | Online Baseball Game | Hockey Simulation | NFL Picks | College Football Picks | Sports Games

© 1999-2014 WhatIfSports.com, Inc. All rights reserved. WhatIfSports is a trademark of WhatIfSports.com, Inc. SimLeague, SimMatchup and iSimNow are trademarks or registered trademarks of Electronic Arts, Inc. Used under license. The names of actual companies and products mentioned herein may be the trademarks of their respective owners.