MLB Series Sweep Math???

**Pancho sanza** · 07-26-11, 12:19 PM

You ain't the first to wonder about this angle.

If anything, I would bet on the team that won the first 2 to complete the sweep, seems folks like to bet against the sweep happening, creating some value on the team going for the sweep.

I've only bet this a handful of times, mostly in the playoffs.

**Carl 1326** · 07-26-11, 01:22 PM

Thanks Pancho! Yes, some folks bet against the sweep while others, like you, bet the streak. Interesting thought as to how the line moves accordingly.
Carl

**TomG** · 07-26-11, 02:59 PM

here is the formula.

pr(team a sweeps a 3 game series) = pr(team a wins game 1) * pr(team a wins game 2) * pr(team a wins game 3)

solved.

extra credit: what assumption does this formula implicitly make?

**ScreaminPain** · 07-26-11, 03:34 PM

Read http://turner.faculty.swau.edu/mathe...rials/playoff/

**illfuuptn** · 07-26-11, 03:50 PM

Originally posted by TomG

here is the formula.

pr(team a sweeps a 3 game series) = pr(team a wins game 1) * pr(team a wins game 2) * pr(team a wins game 3)

solved.

extra credit: what assumption does this formula implicitly make?

It assumes the probabilities are accurate which is the op's question at it's core. He just hasn't reached that level of maturity yet.

It also assumes that "sweep" means "to win all games of a series" as opposed to the traditional definition of the word.

**Carl 1326** · 07-26-11, 05:32 PM

Tom, my question is about the last game of the series after a team has already won 2 or 3 games.

Screamin, thanks for the link! It still dosn't answer answer the question though.

Illf, if the MLB line is 1.50 on the fav, who has won 3 games of a 4 game series, would there be a reason not to bet the fav if you liked them???...

I was hoping someone might have done a study on this.
Carl

**uva3021** · 07-26-11, 06:19 PM

sounds like a job for retrosheet, though it may require some ingenuity for I don't think there is a FATE_SERIES_CT flag

**LT Profits** · 07-26-11, 06:50 PM

Originally posted by Carl 1326

Tom, my question is about the last game of the series after a team has already won 2 or 3 games.

Screamin, thanks for the link! It still dosn't answer answer the question though.

Illf, if the MLB line is 1.50 on the fav, who has won 3 games of a 4 game series, would there be a reason not to bet the fav if you liked them???...

I was hoping someone might have done a study on this.
Carl

In that case, the fact that a team won the first few games of the series is meaningless and you just cap the final game on its own merit and compare your projected line to the real line (just like any other game). The public usually reacts one way or the other to a potential sweep, usually betting on the perceived better team regardless (either going for the sweep or to avoid getting swept by "inferior" team), so there can be value to be found sometimes.

**illfuuptn** · 07-26-11, 07:03 PM

Iow there is no difference in that game's line whether a team is going for the sweep or if the series is tied 1-1. The only reason there would be a difference is if a manager decided to sit some players because they took the first 2 games. And the line would reflect that.

**donjuan** · 07-26-11, 08:06 PM

Originally posted by illfuuptn

It assumes the probabilities are accurate which is the op's question at it's core. He just hasn't reached that level of maturity yet.

It also assumes that "sweep" means "to win all games of a series" as opposed to the traditional definition of the word.

It assumes the games are independent events and not correlated in any way.

**illfuuptn** · 07-26-11, 10:02 PM

^^^the games may be correlated in some ways(such as 2 wins will make the manager sit some players or getting the first two wins used a lot of arms in the bullpen) but the line will reflect that change. Nothing changes about the players themselves whether they won or lost the last few games.

**donjuan** · 07-26-11, 10:35 PM

Talking about the equation not implied market lines.

**Carl 1326** · 07-28-11, 09:12 AM

I think LT hit the nail on the head. Looking for value based on the line being skewed by the "sweep" issue could be the key.
Carl

**rsigley** · 07-28-11, 09:25 AM

people actually know whether or not a team is going to be swept? too much sportscenter IMO

**dbear808** · 07-28-11, 10:06 PM

Through games played before the All-Star break this year when one team has won the first two games of a series of at least three games, the team that can sweep has done so 97 times and has failed 100. I have kept this statistic for many baseball seasons and the percentage has stayed very close to 50-50 during that time. I think a profit can be made by betting against the team that can sweep because you will be betting more teams at +150 than you will be at -175.

A few examples from games played this week:

Seattle @ NY Yankees King Felix help end their horrific losing streak at a price of +140
Colorado @ Los Angeles Colorado avoided the sweep at a price of +135
Houston @ St Louis The horrible Astros beat Carpenter when they were +190
Tampa Bay @ Oakland The Rays were actually a -130 road favorite when they were swept.
NY Mets @ Cincinnati The home-standing Reds could not stop the Mets as a -145 favorite and lost 8-2.

I do not think it is wise to blindly bet all dogs in games where a sweep is involved, just fade the team that can sweep.

**Peregrine Stoop** · 07-29-11, 12:13 AM

Originally posted by Carl 1326

Has anyone done a study of how baseball teams do vs. the MLB odds after winning 2 games in a 3 game series? Or after winning 3 games in a 4 game series? Is it good to bet on the sweep or bet against it? Or maybe it's already factored into the MLB odds? All thoughts welcome.

Carl

you could do the study yourself

**Carl 1326** · 07-29-11, 11:26 AM

Dbear, WOW, thanks! Interesting stuf! I was thinking that there could be value in betting against the sweep.
Carl

**Peregrine Stoop** · 07-29-11, 12:54 PM

Originally posted by dbear808

Through games played before the All-Star break this year when one team has won the first two games of a series of at least three games, the team that can sweep has done so 97 times and has failed 100. I have kept this statistic for many baseball seasons and the percentage has stayed very close to 50-50 during that time. I think a profit can be made by betting against the team that can sweep because you will be betting more teams at +150 than you will be at -175. A few examples from games played this week: Seattle @ NY Yankees King Felix help end their horrific losing streak at a price of +140 Colorado @ Los Angeles Colorado avoided the sweep at a price of +135 Houston @ St Louis The horrible Astros beat Carpenter when they were +190 Tampa Bay @ Oakland The Rays were actually a -130 road favorite when they were swept. NY Mets @ Cincinnati The home-standing Reds could not stop the Mets as a -145 favorite and lost 8-2. I do not think it is wise to blindly bet all dogs in games where a sweep is involved, just fade the team that can sweep.

50/50 is wrong number to be looking at
look at return for betting on team that is going for sweep
look at return for betting on team trying to evade sweep

you could break it up dog v fave, but that would be silly

**Carl 1326** · 07-29-11, 01:24 PM

PS... you got me thinkin. That could be dangerous!

Return on Fav going for a sweep.

Return on Dog going for a sweep. (kinda rare)

That would tell you who to bet, and who to fade... maybe...???
Carl

Oh, but why not separate home and away? Just 4 things.

**Carl 1326** · 07-31-11, 09:49 AM

Lotsa potential sweepers (6) today. Who will sweep and who will bite tha dust??? No big surprise that all are favs.
Carl

**dogman** · 08-01-11, 07:16 AM

Robert Bies has a strategy using the "sweep" principle. He uses the Labby for his preferred way of betting. If anyone wants to check it out, I guess you could google it. He will send you out the plays via e-mail everyday. It is a FREE system you can download with all the rules(he did update the strategy page with a new downloaded version so get the new one) if you want to figure the plays yourself. It is definitely worth it if you're into this type of betting.

I personally don't use it because of the progressive MM he uses. I am not a shill for him just wanted to mention it for those who are into this betting "principle".

**Carl 1326** · 08-01-11, 09:41 AM

Dog, I'm no progressive bettor ether.

The sweeps split out Sunday. If you had bet the fade to risk the same amount you would have cashed since 3 dogs won.
Joe

**Rio DiNero** · 08-01-11, 11:08 AM

I always bet the team to sweep, if they are dogs in every game.

**Carl 1326** · 08-01-11, 02:48 PM

Rio, I'll keep an eye out for those. I am guessing they have won for you.

So maybe just bet the dogs to win in any sweep game situation???

Unless we need some filters?
Joe

**CrimsonQueen** · 08-01-11, 04:01 PM

If you flip a coin 3 times and it has already come up with 3 straight heads, do you think that the fourth flip is still 50/50? The odds of a team completing a sweep are whatever the bookmaker says the odds are on the final game. (within a couple percentage points either way)

**Sawyer** · 08-01-11, 05:05 PM

If road team is going for sweep, usually I try to avoid these kind of games. When motivation is high, stats/numbers can be meaningless..

**Carl 1326** · 08-01-11, 05:36 PM

CQ, not if I have an "all tails" coin palmed. :~) Anyway... I think there could be some value in betting the Dogs here. There or sure was yesterday.

Makes sense to me Sawer! I'll keep an eye on those Homies in these sweep games.

Ok, now I know for sure that a study of dogs/faves/home/away sweepers needs to be done. I just figured somebody had already done it. (^%#@

Oh, I'll bet tha "has" on that prop.

Oh, Dbear has things off to a good start w/ the 50/50 deal.
Carl

Carl

**josie88** · 08-02-11, 12:32 AM

For giggles, toodle over to killersports and input the following queries to see how teams do in the last game of the series:

1) SGS=3 and SG=3 and p:L and pp:L and team (3 game series, lost first two)
2) SGS=3 and SG=3 and p:W and pp:W and team (3 game series, won first two)
3) SGS=4 and SG=4 and p:L and pp:L and ppp:L and team (4 game series, lost first three)
4) SGS=4 and SG=4 and p:W and pp:W and ppp:W and team (4 game series, won first three)

Substitute the words 'conference' and 'division' for 'team' in the queries to see how this plays out for each conference and division.

Some folks also wonder about over/under streaks within a series:

SGS=3 and SG=3 and p:O and pp:O and team (1st two games went over in 3 game series)
SGS=4 and SG=4 and p:O and pp:O and ppp:O and team (4 game series, 3 overs)

Substitute 'U' for 'O' to see the under record and 'conference' and 'division' as above.

Disclaimer: I don't play this angle, but have fun if you do and GL.

**dbear808** · 08-02-11, 02:01 AM

I have prepared a spreadsheet with values through the games of July 20. Stoop's comments prompted me to examine the fav/dog split in more detail. Here are the results:

FAVS 71-39 +16 UNITS
DOGS 37-60 -11 UNITS
TOTAL 108-99 + 5 UNITS

The sheet I use to track each series has 3 rows of 20 spaces to write results in. The next sheet will be complete by Thursday. I will update these results at that time.

To clarify, when the team that can be swept is a favorite in that third game, they avoid the sweep 71 times and are swept 39. When the team that can be swept is a dog in that third game, they avoid the sweep only 37 times and are swept 60.

**Carl 1326** · 08-02-11, 11:28 AM

Josie, I played around at Killersports once but couldn't quite get tha hang of it. Neat that you know how to do it! Any chance of sharing your results?

Dbear, neat! Looks like the dogs are not such a good bet when trying to avoid the sweep. :~( But the favs are and w/ a good ROI.
Joe

**dbear808** · 08-05-11, 11:27 AM

update

Here are updated dog/fav sweep stats:

FAVS: 77-44 +16 UNITS
DOGS: 46-70 -9 UNITS
TOTAL: 123-114 +7 UNITS

**Carl 1326** · 08-06-11, 09:21 AM

Dbear, thanks again! So it looks like betting against the sweep, if the team is danger of being swept is a fav, is a play w/ about a 10% ROI.

Is this the only play?

Playing the favs overall to sweep looks like maybe a break even deal. Maybe by separating favs to sweep by home / away could show another play? You would be betting against the dogs avoiding the sweep that way.
Carl