When trying to get an accurate gauge of how a team (mlb) is performing, is it better to look at 3, 5, or 7 day stats?
3, 5, or 7 day stats
Collapse
X
-
VenkmanSBR Hustler
- 04-16-09
- 85
#13, 5, or 7 day statsTags: None -
uva3021SBR Wise Guy
- 03-01-07
- 537
#2I'd say either 3 or 5 because its within the immediate cycle of the rotationComment -
ljump12SBR High Roller
- 12-08-09
- 113
#4It's not really accurate to look at any of those. Look much longer term to get an idea of how good a team is.Comment -
dwaechteSBR Hall of Famer
- 08-27-07
- 5481
#5What would you possibly think only has relevance over just the past week?Comment -
WrecktangleSBR MVP
- 03-01-09
- 1524
#7Ok, now that I've also done the flip answer, here's my view:
Shorter smoothers tend to work better for shorter season sports (NFL, NCAA FB), and longer smoothers work better for longer season sports (NBA).
Exactly which work better? Actually it seems to differ on the technique. I've even used shorter for some stats and longer for others.Comment -
VenkmanSBR Hustler
- 04-16-09
- 85
#8Thanks for the responses. I asked this because of the streaky nature of teams. Hot streaks versus cold streaks. Should i just forget those split stats and center on pitcher vs batter lifetime stats?
I know there is no one right or wrong way to handicapp a game, I'm just trying to cut out the parts that are wrong. It's my first year handicapping the bases. Just looking to gain some insight.Comment -
sycoogtitSBR Sharp
- 02-11-10
- 322
#9I don't think it's that simple. I would write a program to look at those stats for the past X seasons (where X is large) and see if you can find a trend. Maybe you'll find, for instance, that if a team's last 5 games are WLLWL then there's a 60% chance that they'll lose the next game. I have no idea if that's true or not, I'm just saying you don't know what will work unless you do a lot of backtesting.Comment -
ThrempSBR MVP
- 07-23-07
- 2067
#10Are you seriously trying to draw useful conclusions from anywhere between 10-40 PAs?Comment -
ljump12SBR High Roller
- 12-08-09
- 113
#11Thanks for the responses. I asked this because of the streaky nature of teams. Hot streaks versus cold streaks. Should i just forget those split stats and center on pitcher vs batter lifetime stats?
I know there is no one right or wrong way to handicapp a game, I'm just trying to cut out the parts that are wrong. It's my first year handicapping the bases. Just looking to gain some insight.Comment -
PeepSBR MVP
- 06-23-08
- 2295
#12Streaks certainly happen.
They also happen in coin flips.
Where is Ganchrow when we need him? Would be interesting to know if bases/hoop streaks are of a greater frequency than coin flip streaks (odds adjusted of course).
Anyone know this?Comment -
runnershane14SBR Wise Guy
- 07-23-07
- 803
#13Do players not go into hitting slumps? = streaksComment -
WrecktangleSBR MVP
- 03-01-09
- 1524
#14Ganchrow has died and gone to handicapper's heaven it seems. If it were a result of a natural set of chance events I'd say that streaks don't happen either, but with the human "factor" thrown in, I do think that streaks can occur. Folks can lose faith in themselves.Comment -
ThrempSBR MVP
- 07-23-07
- 2067
#15So what exact criteria would you use over 12 PAs to determine whether someone is on a streak?Comment -
PeepSBR MVP
- 06-23-08
- 2295
#16
I think it is a good question though. What criteria would differeniate a "meaningful streak" from a "coin flipping type" streak?Comment -
PeepSBR MVP
- 06-23-08
- 2295
#17One point that come to mind immediately is you would have to have ways to predict when the streak would start and/or continue.Comment -
wantitall4moiSBR MVP
- 04-17-10
- 3063
#18There are streaks for sure. In basbeall they are a little easier to see and take advantage of since you dont have a spread to worry about.
Spreadbased sports is a little more difficult, but look at any history of a team on a win or losing streak be SU or ATS and watch how those lines get deflated or inflated the longer the streaks go on. So there is an outside force working against streaks in those cases.
Like I said in another thread, dont bet against streak in baseball. Winning streaks or losing streaks. And a streak for the most part starts with the 3rd game. So a team with LL facing a team with WW, is always worth a shot to go with the WW team. Since you have two potential streaks starting.Comment -
ThrempSBR MVP
- 07-23-07
- 2067
#19You're so misinformed its painful.Comment -
duritoSBR Posting Legend
- 07-03-06
- 13173
#20What about 6 days?Comment -
username474SBR Sharp
- 01-09-09
- 480
#21Are you saying that batter vs pitcher stats over 40 PAs don't matter? Are you saying that a .330 batter that is 5 for his last 40 has a better chance of having a productive game then a .270 batter who is 21 for his last 40 ?Comment -
skrtelfanSBR MVP
- 10-09-08
- 1913
#22We should probably consider 4 days too.Comment -
donjuanSBR MVP
- 08-29-07
- 3993
-
ljump12SBR High Roller
- 12-08-09
- 113
#24I haven't done the math -- but i'm going to go with... yes.Comment -
username474SBR Sharp
- 01-09-09
- 480
#25Comment -
wantitall4moiSBR MVP
- 04-17-10
- 3063
#26I love these types of arguments (the batting average one)because usually the math guys just take the numbers at face value. And dont consider the reality of the situation. Nor do they even try to identify subsets.
If a guy has up to a point 2500 ABs and say 340 (13.5%)of them are against 3 pitchers he absolutely OWNS and against those 3 guys his BA is 40 points higher. Thus raising his overall BA a certain amount, I am sure the ath guys CAN figure that out. Then throw in a twist where he cant get a hit of 1 or 2 other guys and his BA against them is 20 points lower than his average. Does that even it out? Or are the 2000 AB against the 'general' pitchers a better determination of his overall ability? If you want to start usuing math you have to ask those questions and get the answers. Especially if their is a clear discrepency.
Then start adding in what ball parks they played in, dome or outside, day or night, how long they were on steroids, how many injuries they have had and so on and so forth.
You guys can cite math all you want, and in some cases it MIGHT work, but more often than not it just confuses the issue. You think if andy Pettite is facing Big Poopy and he says "this guy is 3 for 50 his last 50 AB, but against me he is 30 for 110, he isnt on the juice anymore and it is a day game here in this band box Yankee stadium", so let me get my calculator out to figure out the odds he has of getting a hit off me right now."? Somehow I doubt it.Comment -
ThrempSBR MVP
- 07-23-07
- 2067
#27Actually, that is pretty much almost exactly what you'd do with some higher level math in place of calculators, and using a lot of regression/substitution to league averages for most of the factors as there won't be significant information.
I think its funny a guy who makes his living off people losing money betting gave you five points. Pretty much puts an exclamation point on what you wrote.
How exactly do you figure out how likely he is to get a hit? Guess? Choose a number out of a hat? How do you turn this probability into a betting line? How do you decide what to bet? You're in the wrong forum. And so is Dozer.Comment -
wantitall4moiSBR MVP
- 04-17-10
- 3063
#28Actually, that is pretty much almost exactly what you'd do with some higher level math in place of calculators, and using a lot of regression/substitution to league averages for most of the factors as there won't be significant information.
I think its funny a guy who makes his living off people losing money betting gave you five points. Pretty much puts an exclamation point on what you wrote.
How exactly do you figure out how likely he is to get a hit? Guess? Choose a number out of a hat? How do you turn this probability into a betting line? How do you decide what to bet? You're in the wrong forum. And so is Dozer.
I started with a modest bankroll in 1999 and I pretty much retired from gambling full time 3 years ago, so in 8 years I basically made enough off that bankroll to never have to scramble for a penny at some book that might not ever pay me, or sweat a line move that I am looking for to get an extra few bux out of it ever again. Nor do I have to worry about the government swooping in and arresting me. And had enough money to start a legitimate business.
And I never ever used math beyond what it is best for.
While all the geniuses are broke, gone tout, bonus whoring, or still trying to grind out a few bucks here or there, usually from borrowed money or their last 5 bux to their name.
I can only go by results and empircal evidence. The most valid of which is me, and what I did personally.Comment -
ArilouSBR Sharp
- 07-16-06
- 475
#29wantitall4moi, yes all of those factors should be considered to the extent that the person capping considers them meaningful. If I've been hitting for a while and don't have a few pitchers I hit for different BAs then that would be rather odd in and of itself because random fluctuation should cause 20-40 BA bumps all the time; consider how many ABs a batter gets against each pitcher. Unless you're actually facing one of those pitchers AND you can explain why the effect is there and believe that it is legitimate, the effect should almost entirely cancel.
Yes, doing all of this stuff correctly is hard, cause hey math is hard and statistics is harder. You know what the only thing is that's harder than trying to calculate someone's value factoring in all that stuff using math? Try doing it without using math. The solution to facing a hard problem isn't to pick a number out of thin air or throw up your hands, it is to try and get the best approximation you can and remember that there is an error term. Nothing wrong with a little manual adjustment where you know there's something the equations don't know.
And no, the quite correct answer of Lol yes is not saying that your BA is your BA is your BA, even if BA was the best predictive statistic. You can obviously do vastly better on about five fronts even before situational adjustments. But if you think that this example provided a remotely close calculation, well, let's just say I am very very happy you don't manage (or heaven forbid negotiate trades and contracts for) my ball club. And given who is doing those jobs instead, that's saying something.Comment -
username474SBR Sharp
- 01-09-09
- 480
#30wantitall4moi, yes all of those factors should be considered to the extent that the person capping considers them meaningful. If I've been hitting for a while and don't have a few pitchers I hit for different BAs then that would be rather odd in and of itself because random fluctuation should cause 20-40 BA bumps all the time; consider how many ABs a batter gets against each pitcher. Unless you're actually facing one of those pitchers AND you can explain why the effect is there and believe that it is legitimate, the effect should almost entirely cancel.
Yes, doing all of this stuff correctly is hard, cause hey math is hard and statistics is harder. You know what the only thing is that's harder than trying to calculate someone's value factoring in all that stuff using math? Try doing it without using math. The solution to facing a hard problem isn't to pick a number out of thin air or throw up your hands, it is to try and get the best approximation you can and remember that there is an error term. Nothing wrong with a little manual adjustment where you know there's something the equations don't know.
And no, the quite correct answer of Lol yes is not saying that your BA is your BA is your BA, even if BA was the best predictive statistic. You can obviously do vastly better on about five fronts even before situational adjustments. But if you think that this example provided a remotely close calculation, well, let's just say I am very very happy you don't manage (or heaven forbid negotiate trades and contracts for) my ball club. And given who is doing those jobs instead, that's saying something.Comment -
u21c3f6SBR Wise Guy
- 01-17-09
- 790
#31I happen to use math along with my “concepts” for my selections. Just because I am successful using math for my selections does not mean that someone else cannot be successful using databases, special knowledge, tea leaves or other ways of making selections.
The bottom line to this whole thing regardless of how one comes up with their selections is: Do you collect more on your winning wagers than you lose on your losing wagers (on a flat bet basis)? If you can do that, continue what you are doing regardless of what anyone says. If you are not currently able to do that, consider trying a different approach.
I think sometimes when one is successful “handicapping” a certain way, they lose sight of the fact that there are other ways to “handicap” and be successful as well. Anytime I have attempted to share the “concepts” that I use to make selections to someone that has been gambling for awhile, I usually get a response such as “You can’t do that” or “you have to take this and/or that into consideration” etc. What is even more interesting is when I ask them the question above, “Do you collect more on your winning wagers than you lose on your losing wagers (on a flat bet basis)?”, I generally get, after some dancing, a not exactly answer or something to that effect usually with some excuses thrown in such as “bad beats” etc. The point is, someone was telling me that I couldn’t be profitable doing what I was doing and that I needed to do what they were doing but they were not profitable and I was.
As an example, when I saw the College Basketball semi-finals odds, I could “see” using my “math”, the almost riskless hedge that could be set-up, which I did, and increased my bankroll by a little more than 10% using nothing more than math and my knowledge of the odds. And to top it off, I assume by the standards that some use, both sides of my hedge individually were probably -EV wagers!
As soon as someone says that you can’t use something to help make selections, I am pretty sure that there is probably someone that is doing that successfully. To get this back to the thread question, I don’t know which time-frame is better, if any. That is something you need to investigate because if someone actually had a definitive answer, its use as a tool to find profitable selections will probably be nil as it is probably already accounted for in the odds.
Best of “luck”.
Joe.Comment -
wantitall4moiSBR MVP
- 04-17-10
- 3063
#32Yes, you use EVERYTHING, but dont rely on one thing. Thats why experience is always the best tool a person has.
Math does have its place, dont get me wrong, but more as a way to disprove something rather than prove it. It was like back 10-15 years ago when trends were the big thing, people had trends for everything, and pretty soon those trends started to be called 'probabilities', which is a frigging joke. Just like if the Lakers are now 14-1 (might be 15-1) against the Thunder their odds are 93% to win the game SU. It just doesnt work that way. But that is how most people like to use databases.
None more clear than the -3 in the NFL, everyone says it has a 10% chance (well it used to be like a 12 or 13% chance once) but that percent chance (probability) has changed because fewer games have pushed on the 3 recently so it changed the probabilities. Which is bullshit. But that is the dummy answer.
I have 20 some odd years of NFL data, but since they went to the 2 pt conversion that is really as far back as you need to go. But I took a lot of subsets and found that the games that pushed on 3 that were lined 3 or moved to 3 were mostly accounted for by only 8 teams in the league, and that at one point 3 teams had NEVER ever pushed on the 3 despite having over 100 games combined lined at -3.
In other words there is no widespread probability of ANYTHING happening in sports betting, because the 'probability' is based on past results, not an actual mathematical chance. This isnt cards or dice where the number of possibilities are known, it is sportsbetting where anything and everything can happen. From crooked refs, to kids shaving points, to bad bounces or miracle hoops from half court with no time left on the clock.
So therefore any and all math has to be taken with a shaker of salt not just a grain.
I went to college and grad school with a lot of smart guys, and they were like my own personal ATM machines when it came to sportsbetting. Some of them were really good at cards, but not so much at sports. And there is a reason. You might be able to gues right for alittle while and any variances you might think you have look crystal clear while youre winning, but start losing and all that math goes out the window in a hurry and then it becomes just like every other degenerate out there, chasing good moneyafter bad with desperate plays that youdont have a clue as to why youre playing them.
The bottomline is the best thing any gambler who wants to be successful has to have is the right mentality. No emotions, no favorite teams, no rooting. Not even any cheering or clapping when you win. Hell I didnt even watch sports when I was gambling. Treat it like a business keep the emotions out of it and you will do OK. But if youre just fooling around trying to make a few bucks and want the excitement then worrying about anything too much sort of defeats the whole purpose of that.
But that is what I do these days, fool around and bet a game here or there and watch it on TV. Sort of like if Rembrandt was doing a color by numbers with his grandkids or something.Comment -
WrecktangleSBR MVP
- 03-01-09
- 1524
#33On streaks, here's two examples of there they exist: NFL 2009 last three games of the season for both the Saints and Indy. Both rested players to unknown levels of play or desire to win games. Both teams had performance issues that had to be compensated for during the post season games.Comment -
donjuanSBR MVP
- 08-29-07
- 3993
#34That is pretty blatantly not what OP is talking about.Comment -
MonkeyF0ckerSBR Posting Legend
- 06-12-07
- 12144
#35On streaks, here's two examples of there they exist: NFL 2009 last three games of the season for both the Saints and Indy. Both rested players to unknown levels of play or desire to win games. Both teams had performance issues that had to be compensated for during the post season games.Comment
SBR Contests
Collapse
Top-Rated US Sportsbooks
Collapse
#1 BetMGM
4.8/5 BetMGM Bonus Code
#2 FanDuel
4.8/5 FanDuel Promo Code
#3 Caesars
4.8/5 Caesars Promo Code
#4 DraftKings
4.7/5 DraftKings Promo Code
#5 Fanatics
#6 bet365
4.7/5 bet365 Bonus Code
#7 Hard Rock
4.1/5 Hard Rock Bet Promo Code
#8 BetRivers
4.1/5 BetRivers Bonus Code