Baseball Home Field Advantage

[chubber911]

You can't just try to predict winners in baseball and expect to beat the odds. If you can't put a team on a reasonable winning % going into the game, how do you know if your bet is profitable or not? The biggest mistake in sports betting is the approach where people just try to predict a winner. This is not profitable betting. You need to assign winning %s and compare this against the odds you are laid. How can you expect to assign reasonable percentages without using reasonable stats? Using half of a season for 1 team for your sample size for HFA is not reasonable. They could have played majority against better teams, or majority against worse teams. Fact of the matter is, 40 games (for baseball at least) just isn't enough.

If you made 40 bets, and made money over those bets, do you think anyone would take you seriously if you said you were proven to be a winning handicapper?

[curious]

500 pick challenge

Some crackhead in here challenged me to have a 60% record after 500 picks.

I set up a thread where I will track my next 500 picks. I don't flat bet so I am not accepting the rule that all plays are for 1 unit. I'm also not accepting the rule that I use closing lines. I'll use the line I got when I posted the picks.

When the 500 plays are over I'll come back here and post the results.

I'm not interested in arguing about minutiae about this. I will record what happens in a public forum for all to see. At the end of the 500 picks I will be up > 60% or not.

Since I don't flat bet, the crackhead will have to tell me if my record after 500 plays meets his criteria or not.

[curious]

You can't just try to predict winners in baseball and expect to beat the odds. If you can't put a team on a reasonable winning % going into the game, how do you know if your bet is profitable or not? The biggest mistake in sports betting is the approach where people just try to predict a winner. This is not profitable betting. You need to assign winning %s and compare this against the odds you are laid. How can you expect to assign reasonable percentages without using reasonable stats? Using half of a season for 1 team for your sample size for HFA is not reasonable. They could have played majority against better teams, or majority against worse teams. Fact of the matter is, 40 games (for baseball at least) just isn't enough.

If you made 40 bets, and made money over those bets, do you think anyone would take you seriously if you said you were proven to be a winning handicapper?

I use winning % for all my predictions which is why I don't flat bet.

I use strength of schedule in my weightings for home/away.

I'm not sure where you came up with your assumptions, but I don't know anyone who does this seriously who does not assign probabilities to the outcomes and does not use a strength of schedule weighting.

[chubber911]

I'm not sure where you came up with your assumptions, but I don't know anyone who does this seriously who does not assign probabilities to the outcomes and does not use a strength of schedule weighting.

You're arguing that I'm wrong, but then your point is saying serious bettors do what I described in my post? Or are you arguing that everyone that takes betting seriously knows what they're doing?

Or maybe I'm just confused. What assumption of mine are you saying is incorrect?

[coldhardfacts]

You're arguing that I'm wrong, but then your point is saying serious bettors do what I described in my post? Or are you arguing that everyone that takes betting seriously knows what they're doing?

Or maybe I'm just confused. What assumption of mine are you saying is incorrect?

I'm a serious bettor. I bet on baseball. I win consistently year after year after year.

Home field advantage/road field disadvantage relative to specific teams is not the only factor or even the most important factor used in handicapping games. But it is an important factor. Anyone who thinks every team plays relatively as well at home as they do on the road is DEAD WRONG and has no understanding of baseball. And I will guarantee you, if this is how they factor home field advantage in handicapping they are LOSERS. If you're saying that 40 games isn't a reasonable sample size, then I guess all the batting averages, ERAs, slugging pctgs, etc. etc. etc. year to date are pretty meaningless, also.

[coldhardfacts]

If you don't know anything about statistics, you don't know anything about baseball.

What an idiotic statement. Baseball and statistics are not synonomous. You can know something about statistics and nothing about baseball, and you can know something about baseball and nothing about statistics. How many stat courses did Babe Ruth take?

[chubber911]

I said a team's record, alone, in 40 games isn't enough to quantify their home field advantage.

[curious]

I said a team's record, alone, in 40 games isn't enough to quantify their home field advantage.

Which is dead wrong.

[curious]

You're arguing that I'm wrong, but then your point is saying serious bettors do what I described in my post? Or are you arguing that everyone that takes betting seriously knows what they're doing?

Or maybe I'm just confused. What assumption of mine are you saying is incorrect?

You assume that you are talking to people who do not know how to cap baseball because you make these assumptions about us:

You can't just try to predict winners in baseball and expect to beat the odds. If you can't put a team on a reasonable winning % going into the game, how do you know if your bet is profitable or not? The biggest mistake in sports betting is the approach where people just try to predict a winner. This is not profitable betting. You need to assign winning %s and compare this against the odds you are laid.

Every serious gambler I know assigns probabilities.

Using half of a season for 1 team for your sample size for HFA is not reasonable. They could have played majority against better teams, or majority against worse teams. Fact of the matter is, 40 games (for baseball at least) just isn't enough.

Every serious gambler I know uses strength of schedule. 40 games is certainly enough to determine trends. If it isn't then you are saying that there is no point in looking at batting average or ERA or walk to strikeout ratio in April or May which is an idiotic statement to make.

[donjuan]

Curious,

2 things:

1. Show your work as Joe asked.
2. Picking >60% is meaningless without talking about a specific price or narrow range of prices. Anyone can hit >60% if that is their goal.

[chubber911]

Curious, now it just sounds like you are purposefully trying to be antagonistic. Do you think every person that reads this thread knows as much about sports betting as you say you know? Why are you so eager to discredit me by commenting about ERA and other pitching stats, when this thread was about determining teams home field advantages? It's not enough that you have to be right, but anyone you think is opposing your opinion must be proven as wrong.

I thought someone that knows as much as you would also know that different stats converge at different rates. Why are you now the second person saying that a team's winning percentage is just as close to a true value after 40 games as individual player stats?

Judging by the tone you take in your responses, it appears to me that you are actually the one who assumes you are talking to people that know nothing about handicapping. But then again, I am idiotic.

[coldhardfacts]

Curious, now it just sounds like you are purposefully trying to be antagonistic. Do you think every person that reads this thread knows as much about sports betting as you say you know? Why are you so eager to discredit me by commenting about ERA and other pitching stats, when this thread was about determining teams home field advantages? It's not enough that you have to be right, but anyone you think is opposing your opinion must be proven as wrong.

I thought someone that knows as much as you would also know that different stats converge at different rates. Why are you now the second person saying that a team's winning percentage is just as close to a true value after 40 games as individual player stats?

Judging by the tone you take in your responses, it appears to me that you are actually the one who assumes you are talking to people that know nothing about handicapping. But then again, I am idiotic.

If you are implying that I am the first to say that a team's winning percentage is JUST AS CLOSE TO A TRUE VALUE AFTER GAMES AS INDIVIDUAL PLAYER STATS, then you better take a reading comprehension course, and yes, you are being idiotic. I never said any such thing. (In fact I said the opposite.)

The fact is, a team's record through 40 games is indicative of how well they play at home or on the road. Have you ever heard of Earl Weaver? Do you think he was a little bit successful? He would make lineup/pinch hitting/relief pitcher decisions based on statistical matchups of as few as 5 or 6 at bats, or sometimes even fewer. It's a good thing he knew a little bit more about baseball than the stat "geniuses" on this forum who insist that sample size is the be all and end all.

[curious]

Curious,

2 things:

1. Show your work as Joe asked.
2. Picking >60% is meaningless without talking about a specific price or narrow range of prices. Anyone can hit >60% if that is their goal.

I am showing it, I started a 500 pick thread in the baseball forum as I said. I will update it daily. I will play all my plays as I normally would. I play a variety of lines. I don't flat bet. After the 500 plays you experts can say if I met your criteria or not. if at that point I am > 60% and you think that is "meaningless" well that will be your opinion.

I have nothing more to say regarding this matter. I will talk about it from now on with my record as captured in my 500 pick thread.

[curious]

Curious, now it just sounds like you are purposefully trying to be antagonistic.

Actually, you implied that we did not know enough about capping to use strength of schedule and probability of W/L instead of just W/L. Which I thought was pretty insulting. Any capper who knows anything knows to use strength of schedule and win probability.

Do you think every person that reads this thread knows as much about sports betting as you say you know?

I wasn't talking about everyone who reads the thread, I was talking about the people you were arguing with.

Why are you so eager to discredit me by commenting about ERA and other pitching stats, when this thread was about determining teams home field advantages? It's not enough that you have to be right, but anyone you think is opposing your opinion must be proven as wrong.

I was illustrating that your comment that 40 games is too small a sample size is not supported by the facts. We use lots of stats that have a much smaller sample size than 40 games. I just picked ERA as an example, I could have picked bases per run or slugging or any other stat just as well.

I thought someone that knows as much as you would also know that different stats converge at different rates. Why are you now the second person saying that a team's winning percentage is just as close to a true value after 40 games as individual player stats?

I wasn't talking about winning %, I was talking about HFA which takes into consideration more variables than just winning %. Well, you are saying that HFA stats for the entire league are meaningful, I say they are not, that you have to use HFA for specific teams. I don't agree that the convergence of league stats for HFA is meaningful. So, you insist on a very large sample size because of your premise. I say the premise is wrong. To be honse, I'm not really sure what this means "different stats converge at different rates". Which stats? Give some examples.

Judging by the tone you take in your responses, it appears to me that you are actually the one who assumes you are talking to people that know nothing about handicapping. But then again, I am idiotic.

Nope. I was reponding to your talking to us like we are total novices who don't know what strength of schedule is and don't know that Win probability is much more important than simple W/L.

Ask anyone who knows me, my tone in this thread is actually quite polite compared to my normal volatile behavior.

[donjuan]

I am showing it, I started a 500 pick thread in the baseball forum as I said. I will update it daily. I will play all my plays as I normally would. I play a variety of lines. I don't flat bet. After the 500 plays you experts can say if I met your criteria or not. if at that point I am > 60% and you think that is "meaningless" well that will be your opinion.

I have nothing more to say regarding this matter. I will talk about it from now on with my record as captured in my 500 pick thread.

If you can't comprehend that claiming to win x% of your games without a specific price or narrow price range attached is completely meaningless, then you'd probably find Players Talk more suitable. Picking a bunch of -220 favs in baseball is a surefire way to win more than 60% of your games. It should also be pointed out that using 500 bets at non-widely available lines does nothing to prove your HFA-related assertion earlier in the thread.

[chubber911]

I'm sorry if I sounded condescending with my initial response. It was more of a general observation (one which obviously veteran cappers comprehend), in regards to individuals of the mind set "oh, Jays never win at New Yankee Stadium".

I'm still trying to learn how to quantify HFA, myself, and I'm somewhere in between assigning a generic value and weighting the specific criteria. Initially, I had team winning % on my mind when I first entered this discussion, and not necessarily how HFA changes batting averages, WHIP, etc.

Lastly, you asked if I could explain and give examples about what I meant when I said "different stats converge at different rates". Please don't take my explanation as an insult, I'm just being thorough with this answer so there is no confusion. What I mean is: after x games, certain statistics are going to be much more reliable than others. After 10 games, there are probably several players whose batting averages aren't very close to their true estimation of a batting average. A batter could have been very hot and be hitting 20 for 45 at that point, but I don't think that means you can expect them to be hitting over .400 going into the next game. A pitcher could have a WHIP of 1.00 and an ERA of 6.50 after 3 games, and you'd probably agree that trying to quantify this pitchers ERA as such when capping a game isn't very accurate. That's what I meant when I said that quote. Sorry if this is obvious and I didn't answer your question.

I've been a poker player for years and play professionally, so I'm a person that loves to see very specific stats, but I'm also cynical about stats where I don't think I have enough data to be completely accurate. I have only been capping sports for less than a year, so there's lots I need to learn, but for better or worse, my habits from poker carry over to handicapping.

[coldhardfacts]

I'm sorry if I sounded condescending with my initial response. It was more of a general observation (one which obviously veteran cappers comprehend), in regards to individuals of the mind set "oh, Jays never win at New Yankee Stadium".

I'm still trying to learn how to quantify HFA, myself, and I'm somewhere in between assigning a generic value and weighting the specific criteria. Initially, I had team winning % on my mind when I first entered this discussion, and not necessarily how HFA changes batting averages, WHIP, etc.

Lastly, you asked if I could explain and give examples about what I meant when I said "different stats converge at different rates". Please don't take my explanation as an insult, I'm just being thorough with this answer so there is no confusion. What I mean is: after x games, certain statistics are going to be much more reliable than others. After 10 games, there are probably several players whose batting averages aren't very close to their true estimation of a batting average. A batter could have been very hot and be hitting 20 for 45 at that point, but I don't think that means you can expect them to be hitting over .400 going into the next game. A pitcher could have a WHIP of 1.00 and an ERA of 6.50 after 3 games, and you'd probably agree that trying to quantify this pitchers ERA as such when capping a game isn't very accurate. That's what I meant when I said that quote. Sorry if this is obvious and I didn't answer your question.

I've been a poker player for years and play professionally, so I'm a person that loves to see very specific stats, but I'm also cynical about stats where I don't think I have enough data to be completely accurate. I have only been capping sports for less than a year, so there's lots I need to learn, but for better or worse, my habits from poker carry over to handicapping.

I'm also a poker player, although I'm still struggling to achieve the consistent success there that I have through sports betting. I think my shortfall is that poker requires more relatively instantaneous decision-making than sports, where you have unlimited time to analyze and evaluate specific plays. Anyway, think of it this way: From the time you sit down at a table, you're trying to draw a read on the other players and how they play. The statistical basis for your analysis is very limited - you are observing only an infitesimal percentage of the total hands they have played. Still, you must size them up relatively quickly in order to determine how loose or tight they are, and how they are likely to play certain game situations based on their tendencies. And while this observation is only one of many factors that determines how you're going to play any hand, it is an important factor nonetheless.

[arwar]

14468 games homefield results

I took the spreadsheet that someone referenced and extracted:

HOMETEAM W/L
ML= -200 #of games= 27 total=$ -300 grand_total =$- 300
ML= -195 #of games= 4 total=$ -190 grand_total =$- 490
ML= -190 #of games= 24 total=$-1370 grand_total =$- 1860
ML= -185 #of games= 83 total=$ -250 grand_total =$- 2110
ML= -170 #of games= 158 total=$ -400 grand_total =$- 2510
ML= -165 #of games= 146 total=$ 820 grand_total =$- 1690
ML= -160 #of games= 211 total=$ 820 grand_total =$- 870
ML= -155 #of games= 238 total=$-2210 grand_total =$- 3080
ML= -150 #of games= 239 total=$ 150 grand_total =$- 2930
ML= -145 #of games= 278 total=$-2580 grand_total =$- 5510
ML= -140 #of games= 383 total=$-2260 grand_total =$- 7770
ML= -135 #of games= 355 total=$ 720 grand_total =$- 7050
ML= -130 #of games= 476 total=$-1850 grand_total =$- 8900
ML= -125 #of games= 398 total=$-1375 grand_total =$-10275
ML= -120 #of games= 532 total=$-1140 grand_total =$-11415
ML= -115 #of games= 527 total=$-2125 grand_total =$-13540
ML= -110 #of games= 613 total=$-1910 grand_total =$-15450
ML= -105 #of games= 523 total=$-1410 grand_total =$-16860
ML= -100 #of games= 2 total=$ 200 grand_total =$-16660
ML= +100 #of games= 710 total=$-1200 grand_total =$-17860
ML= +105 #of games= 685 total=$-2490 grand_total =$-20350
ML= +110 #of games= 659 total=$-4160 grand_total =$-24510
ML= +115 #of games= 587 total=$ -5 grand_total =$-24515
ML= +120 #of games= 744 total=$-4440 grand_total =$-28955
ML= +125 #of games= 605 total=$ -875 grand_total =$-29830
ML= +130 #of games= 675 total=$-1490 grand_total =$-31320
ML= +135 #of games= 584 total=$ 4580 grand_total =$-26740
ML= +140 #of games= 528 total=$ 1440 grand_total =$-25300
ML= +145 #of games= 536 total=$ 790 grand_total =$-24510
ML= +150 #of games= 520 total=$-1000 grand_total =$-25510
ML= +155 #of games= 435 total=$-1935 grand_total =$-27445
ML= +160 #of games= 354 total=$ -820 grand_total =$-28265
ML= +165 #of games= 302 total=$ -255 grand_total =$-28520
ML= +170 #of games= 271 total=$-1180 grand_total =$-29700
ML= +175 #of games= 235 total=$ 1250 grand_total =$-28450
ML= +180 #of games= 270 total=$-1800 grand_total =$-30250
ML= +185 #of games= 146 total=$-1205 grand_total =$-31455
ML= +190 #of games= 117 total=$-1260 grand_total =$-32715
ML= +195 #of games= 54 total=$-1860 grand_total =$-34575
ML= +200 #of games= 234 total=$-1200 grand_total =$-35775

I have another sheet listing all the dates, games, scores,
and lines. I used the 10cent line

[skrtelfan]

Baseball Prospectus recently had an interesting 4 or 5 part series on HFA written by Matt Swartz. You need to be a paid subscriber to read it but their subscriptions are cheap, only $40 a year I think.

[evo34]

Baseball Prospectus recently had an interesting 4 or 5 part series on HFA written by Matt Swartz. You need to be a paid subscriber to read it but their subscriptions are cheap, only $40 a year I think.

Yeah, that was really good stuff. BP has gone way downhill in the past 2-3 years, but Swartz is a great addition.

His article here, http://www.thegoodphight.com/2008/7/...ield-advantage, trounces the argument that team-specific HFA is predictive, even using a whole season of data. It's awesome that some people (coldhardfacts and curious) insist on using 40 games of data to predict a team's rest-of-season HFA. These are the people winning bettors need to exist in order to make money.