How crazy is the current hot streak for favorites in the UFC?

[Svino]

Pretty damn crazy.

• If you bet just one unit on every favorite, you would be +10.4u for the year so far.
• Dogs higher than +200 have gone 1-19, with the only win being Prater's DQ victory over Erick Silva.
• Even if we neglect the vig, the odds that favorites overall would do this well (or better) by random chance is about 1 in 55.
• With a simple model for removing the vig, the odds are more like 1 in 200.
• A person who bet every favorite in my contest would be winning with a score of 5.98.


To make comparisons to past years in the UFC, I have tabulated the average return on a 1-unit fixed-swing bet for different line ranges. With BookMaker's vig (I am using their closing lines), we would expect this value to be around -0.025 for a random bet.

2010
-116 to -199 = -0.088
-200 to -249 = -0.051
-300 to -399 = -0.047
-400 to -599 = +0.074
-600 and lower = -0.155
All favorites: -0.060
All dogs: +0.010

2011
-116 to -199 = -0.101
-200 to -249 = +0.040
-300 to -399 = +0.058
-400 to -599 = +0.044
-600 and lower = -0.238
All favorites: -0.018
All dogs: -0.031

2012 so far
All favorites: +0.140
All dogs: -0.199

• It's interesting to see that 2010, a year in which I formed many of my betting habits, was actually a nice year for dogs. A blind 1u bet on each underdog would have left you with +5.5u for the year.

• 2011 was definitely a better year for favorites than 2011, with value overall in the -200 to -600 range.

• There is obviously a lot of fluctuation with the values here, and possible analysis pitfalls with arbitrary binning / cherrypicking data and such. That said, I am curious if there is some staying power to the idea that -400 to -600 favorites are good bets, and favorites lighter than -200 are weak.

• It doesn't even take that big an edge to be profitable. I had very good years in 2010 and 2011 with average returns on lines I bet of +0.064, and+0.051 (2012 so far: -0.032). The +0.14 return on all favorites so far is insane.

• I still don't know what to make of all this. I don't know if there is something here that represents a change in the way the UFC, or bettors, or oddsmakers are doing things, or if it really is just chance. I am sure it can't be sustainable; bettors will start to change their plays if nothing else.

[MMA_scientist]

Interesting, thanks for doing the work.

I have been saying for a couple weeks that the lines just seem different to me, and I have actually been betting dogs a lot (which I have rarely done in the past). I think the lines makers might be adjusting the lines, making them a bit steeper. So far, I have to think it is just chance... there have been seveat least a few big dogs (Bisping) that could have been given decisions. Also, it seems like the UFC is making fewer mismatches lately, Oliveira/Wisely and Palhares/Massenzio are the only out and out squash matches I can think of so far in 2012 (there may be others though).

I don't know, I hope it adjusts soon. I really don't want to see MMA lines end up looking like boxing lines (seems like it is common to see -1000 favorites every week in pretty major fights). Maybe the lines are sharper in boxing and the MMA gravy train is drying up... I think we will still figure out a way to beat the book, but it has been pretty damn easy since I started betting a few years ago.

[SPX]

Good work there, detective!

I know it's obvious, but I really think the key to being profitable with betting large favorites in the -300 to -500 range (-500 being my personal limit in ALMOST all cases) is to be selective. That is, make sure the line is justified and is not based on hype or that some legitimate way to win on the underdog's part is not being overlooked. So basically, does the dog really, truly have almost no way to win the fight?

As long as you are good at making that assessment you should win more than you lose.

[SPX]

BTW, Jay Silva choking out Grove is a perfect example of some shit that I would just never think would really happen.

[Ludo]

Well Grove does kind of suck.

[SPX]

He's actually a really competent fighter and borderlined UFC calibur.

[Ludo]

He's actually a really competent fighter and borderlined UFC calibur.

He's very hit or miss, though. His ability to take a shot in the most striker heavy division in the sport is astoundingly bad and he still hasn't filled into his frame yet(suggesting he probably won't ever do so). His length and height have allowed him to bridge certain skill gaps that he otherwise wouldn't have been able to overcome.

[SPX]

I agree he's hit or miss, but I would expect him to take care of someone like Jay Silva pretty much every time out.

[Luke]

He's actually a really competent fighter and borderlined UFC calibur.


BOL::lmao::


I guess you missed his stellar performance against Minowa

[SPX]

I did see him clearly whip Minowa's ass.

[Luke]

I did see him clearly whip Minowa's ass.


BOL,he arguably lost 1 rd.(or was that rd 2?)

Not to mention Minowa is a bum x2

[Ludo]

BOL,he arguably lost 1 rd.(or was that rd 2?)

Not to mention Minowa is a bum x2

All three judges had it 30-27 for Grove. Thats not losing a round.

[Luke]

All three judges had it 30-27 for Grove. Thats not losing a round.


Are you fukin stupid? I said arguably

[Ludo]

Are you fukin stupid? I said arguably

No, I was trolling you in the same sense you trolled X about the Machida/Jones fight. Way to keep a lid on it, hoss.

[Luke]

Thank you for deducting reputation from this user

......................

[Ludo]

So much butthurt..............

[SPX]

BOL,he arguably lost 1 rd.(or was that rd 2?)

Not to mention Minowa is a bum x2

What's with the Minowa hate? That dude's a boss. . .

[SPX]

So much butthurt..............

BOL!

[Svino]

That is, make sure the line is justified and is not based on hype or that some legitimate way to win on the underdog's part is not being overlooked. So basically, does the fave really, truly have almost no way to win the fight?

One rule I try to follow is to imagine that the favorite is just a little worse in all areas than you think he is, and the underdog is just a little better. If that would give the underdog a decent chance, stay away from steep lines.

For fun: here are all the -500 or steeper losers from 2009 to present:

BJ Penn (-1100) vs. Frankie Edgar
Here, the fact that Edgar's "hummingbird style" was perfect for the type of outside striking match this fight was likely to be should have kept people away. Even so, it took a bad decision.

Rolles Gracie (-1010) vs. Joey Beltran
UFC newcomers = too much uncertainty.

Ryan Bader (-625) vs. Tito Ortiz
I'm not 100% convinced this line was bad. Maybe Bader just got caught?

Tom Lawlor (-620) vs. Joe Doerksen
The definition of a "grappler's chance".

Melvin Guillard (-615) vs. Joe Lauzon
Guillard rightly deserved to be a big favorite, but I guess if a fighter is undisciplined, there's a limit to how steep you should bet.

Erick Silva (-600) vs. Carlo Prater
Freak DQ, probably a good line, given the beating Silva put on him.

Josh Koscheck (-575) vs. Paulo Thiago
Good old K1 Kos. Wrestler who is too willing to stand vs. a guy with good power. Didn't scare me away from taking Kos at pretty heavy odds against Daley, though.

John Hathaway (-525) vs. Mike Pyle
This is the only one I actually bet (on Pyle).

Rick Story (-525) vs. Charlie Brenneman
This is a case where the rule I stated above would apply. People assumed Story would either outwrestle Brenneman, or their wrestling would cancel.

Brock Larson (-500) vs. Brian Foster
Brock Larson (-500) vs. Mike Pierce
Larson was -500 for three fights in 2009 and he had the good fortune to win one of them. But he didn't just lose both of these fights, he got dominated. LOL Larson.

[MMA_scientist]

don't you love it when you put in a bunch of work, post some new kind of stuff- get no replies- and the 15 posts arguing about minowa. IWS sucks so much, I am going to zewkey, catch you guys on the flipflop.

[Svino]

I got the UFC 2009 stats, too. It did also have the pattern of a positive -400 to -600 range and bad -116 to -119 range.

2009
-116 to -199 = -0.045
-200 to -299 = +0.044
-300 to -399 = -0.175
-400 to -599 = +0.030
-600 and lower = +0.120
All favorites: -0.028
All dogs: -0.018

2009 - present
-116 to -199 = -0.064
-200 to -299 = +0.013
-300 to -399 = -0.042
-400 to -599 = +0.053
-600 and lower = -0.142
All favorites: -0.023
All dogs: -0.025


The more I think about the current hot streak for favorites, the more I'm sure it must be coincidence. There have been other streaks, just not as big, and if you were to flip some decisions like Koscheck, Figueroa, etc., it wouldn't look that much out of normal. It will be business as usual for me unless this continues for a few more events.

[Svino]

don't you love it when you put in a bunch of work, post some new kind of stuff- get no replies- and the 15 posts arguing about minowa.

LOL. I have always have a soft spot for the conspiracy theory that Minowa is secretly the best fighter ever and all his losses are works.

[MMA_scientist]

My theory is that Minowa is the worst fighter in the world, and that all of his wins are works. All evidence points to the super hulk tournament being totally fixed.

[SPX]

One rule I try to follow is to imagine that the favorite is just a little worse in all areas than you think he is, and the underdog is just a little better. If that would give the underdog a decent chance, stay away from steep lines.

I think you pretty much nailed it there.



Ryan Bader (-625) vs. Tito Ortiz
I'm not 100% convinced this line was bad. Maybe Bader just got caught?

The line probably wasn't too far off. I mean, can you really see Tito doing it again?

Poor Tito. At least he did get one more good win before closing out his career.



Brock Larson (-500) vs. Brian Foster
Brock Larson (-500) vs. Mike Pierce
Larson was -500 for three fights in 2009 and he had the good fortune to win one of them. But he didn't just lose both of these fights, he got dominated. LOL Larson.

BOL.

I think the problem here is that he was fighting a couple of guys who were being being severely underestimated. We know now that both Pierce and Foster are pretty beastly.



don't you love it when you put in a bunch of work, post some new kind of stuff- get no replies- and the 15 posts arguing about minowa. IWS sucks so much, I am going to zewkey, catch you guys on the flipflop.

Hey, well I did make a legit reply. But I know what you're saying.

I remember when Ludo (I think) posted that UFC Champs By the Numbers thing or whatever it was called. I thought it was really interesting but apparently no one wanted to talk about it.

[MMA_scientist]

[quoute=Svino]The more I think about the current hot streak for favorites, the more I'm sure it must be coincidence. There have been other streaks, just not as big, and if you were to flip some decisions like Koscheck, Figueroa, etc., it wouldn't look that much out of normal. It will be business as usual for me unless this continues for a few more events. [/quote]

That's what I'm saying. I think it is just an usually long run, but that is bound to happen from time to time with -600 favorites.

[MMA_scientist]

I would have commented on the Champs by the Numbers thing, but I had already seen it on SD and its moment had passed in my mind

[Svino]

Poor Tito. At least he did get one more good win before closing out his career.

That would have been the perfect time for him to retire. But these guys always get their late career win and then they're like, "See! I've still got it! Another run at the title!" Now we have to watch him get beat up by Griffin and close with a 3-fight losing streak.

[SPX]

I would have commented on the Champs by the Numbers thing, but I had already seen it on SD and its moment had passed in my mind

If you're not part of the solution, then you're part of the problem.

[SPX]

BTW Svino, all this stuff seriously is interesting. Thanks for doing the work.

I do wonder how the results would be different though if you picked the best line, or the average line, instead of the closing lines. I mean, I know I bet a lot of these guys at lines that were much better than what you're using.

[Luke]

This shits too much like Stats class.............going to zewkey also.

[Luke]

Where are you finding all these lines at?

Also,wouldn't using opening lines instead of closing lines give us more info since it's before the public moves the line?

[Svino]

BTW Svino, all this stuff seriously is interesting. Thanks for doing the work.

I do wonder how the results would be different though if you picked the best line, or the average line, instead of the closing lines. I mean, I know I bet a lot of these guys at lines that were much better than what you're using.

Yeah, I did too. Sometimes the best line is too fleeting to be relevant to me though, even with BFO sending me alerts. One thing I would like to look at someday is if there's any pattern with the relationship between the opening and closing lines. Like: for fighters that close at -200, do guys that opened at -115 do better or worse than guys that opened at -330?

[SPX]

One thing I would like to look at someday is if there's any pattern with the relationship between the opening and closing lines. Like: for fighters that close at -200, do guys that opened at -115 do better or worse than guys that opened at -330?

I'd certainly be curious to know. Get on it!

[Svino]

Where are you finding all these lines at?

Also,wouldn't using opening lines instead of closing lines give us more info since it's before the public moves the line?

I am using the closing BookMaker lines from bestfightodds.com. Using opening lines would certainly give *different* info. But what I'd ideally like is to use the line that I would be able to, or most likely to bet the fight at. For me, that's probably closer to the closing line. Not only do opening lines often not last longer than about 15 minutes, the books sometimes only accept small bets on them.

[MMA_scientist]

That would be interesting and useful information to have. Do that. I will wait.

[SPX]

For me, that's probably closer to the closing line. Not only do opening lines often not last longer than about 15 minutes, the books sometimes only accept small bets on them.

Why not use an average line then? I mean, a lot of times lines don't move substantially until the day of the fight, and at that point we've often already made our bets. I'm positive Penn was under -500 for at least a little while--long enough to get a bet in--before jumping up to -1100 (which I didn't even know ever happened).

Guillard also was under -500 for a while, at least long enough for any observant person to place their bet.

[Svino]

Why not use an average line then?

Unfortunately, I'm not sure the data needed to take a proper time average is available anywhere. The data needed to do it "improperly" might be available on BFO, but it would be very time consuming to get. Hmm. I'd have to figure out a way to write a script to get it from their pop-up thing.

[Ludo]

Awesome work, Svino. I'm a little bit of a stat junkie when it comes to some things in MMA but I never really looked at winning percentages vs steep lines, I only started making the connection that there was a trend in favorites winning more frequently at the end of last year when My record on favorites was literally the exact opposite and double My record on dogs(which was what I made My money on when I got into betting the sport initially).

Minowa is probably the only legend of the sport for being involved in more suspicious fights than any other figure to date.

I'm not even bothered by the fact that nobody wanted to talk about the champions by the numbers thing. I realize not everyone is a stat junkie.

[Svino]

OK, one more stats thing. This is basically a continuously-blurred (instead of binned discretely) version of the 2009-to-present data. The average return on a fixed-swing bet is plotted against the implied win probability of the line (with vig). The data can't be trusted too much out on the edges, where there are few fights, but the positive peaks in the +120 to +100 range, and the peak a little over and around -400 stand out.http://img528.imageshack.us/img528/4218/2009presentwblur.png

[Ludo]

OK, one more stats thing. This is basically a continuously-blurred (instead of binned discretely) version of the 2009-to-present data. The average return on a fixed-swing bet is plotted against the implied win probability of the line (with vig). The data can't be trusted too much out on the edges, where there are few fights, but the positive peaks in the +120 to +100 range, and the peak a little over and around -400 stand out.http://img528.imageshack.us/img528/4218/2009presentwblur.png

I'm going to assume I'm not alone when I say that I have no fucking idea what that graph is trying to tell Me.

[SPX]

I for damn sure don't.

[Svino]

I'm going to assume I'm not alone when I say that I have no fucking idea what that graph is trying to tell Me.

Sorry. For example, the height of the line at (say) 80, would give you an estimate of expected return for bets on fighters around -400 (since that corresponds to implied odds of 80%). Places where the line is above zero indicate lines or line ranges where blind-betting on all fighters in that range from 2009 to the present would have been profitable.

[Ludo]

Sorry. For example, the height of the line at (say) 80, would give you an estimate of expected return for bets on fighters around -400 (since that corresponds to implied odds of 80%). Places where the line is above zero indicate lines or line ranges where blind-betting on all fighters in that range from 2009 to the present would have been profitable.

Thanks, I think I have it now.

[SPX]

My theory is that Minowa is the worst fighter in the world, and that all of his wins are works. All evidence points to the super hulk tournament being totally fixed.

What evidence exactly are you referring to? The evidence of Minowa winning?

[Ludo]

What evidence exactly are you referring to? The evidence of Minowa winning?

Plenty of fighters have come out and said they were offered money to take dives against Minowa in the past. He's like the Goldberg of Japan.

[SPX]

Plenty of fighters have come out and said they were offered money to take dives against Minowa in the past. He's like the Goldberg of Japan.

I don't really follow his career so I haven't heard that. But it's interesting.

[Mr. IWS]

So much butthurt..............

Bulz

[Mr. IWS]

BJ Penn (-1100) vs. Frankie Edgar
Here, the fact that Edgar's "hummingbird style" was perfect for the type of outside striking match this fight was likely to be should have kept people away. Even so, it took a bad decision.

Rolles Gracie (-1010) vs. Joey Beltran
UFC newcomers = too much uncertainty.

Melvin Guillard (-615) vs. Joe Lauzon
Guillard rightly deserved to be a big favorite, but I guess if a fighter is undisciplined, there's a limit to how steep you should bet.




Srs nigga, you cut me deep with this shit.

We are not friends anymore.

[Svino]

2012 so far
All favorites: +0.140
All dogs: -0.199


... that was two events ago. And after two events in which 15 of 23 underdogs hit, we have the much more reasonable-looking:

2012 so far
All favorites: +0.004
All dogs: -0.053

So I guess things sorted themselves out pretty quick after all that.

[MMA_scientist]

Yeah, I had a feeling this was going to balance out eventually. It is pretty crazy how it works out almost perfectly though.