i feel the same way about Frankie but its hard to pick against a guy who's so scrappy & has very good fight IQ. to me its a pickem, my gut says Yair and my head says Frankie

I already got Yair at 2.5, but I think it's fair that Frankie is the favorite. He has not declined that much IMO, he still got a great chin, and is though, speed is still fine IMO.

Frankie did great against Swanson another creative unpredictable striker and controlled him on the ground for most of the fight. Franky is great at catching kicks, and Yair throws a lot. So we could see something similar. Or maybe Yair just gets the KO if he lands one of his headkicks he indeed could also outworks him and win a decision.

Should be a GREAT fight, a fan of both fighters, might just hedge my bet and take a small guaranteed win, and enjoy the fight. UFC 211 is such a great card!

Well, if business is business (which I agree with), then TJ, TAM best fighter at the time, leaving for a competing gym is a harsh blow which demands some sort of reaction.
Now, one course of action is to put forward a dilemma to fighters who are weighing the option of trying out a different gym, making the tryout not risk free. Faber set out to use TJ as an example, naturally.

There cannot be any exchange between the two different gyms as they are competing against each other. In that sense, TJ was rather naive believing he could represent Duane Ludwigs gym while at the same time training at TAM gym. I believe TJ made the correct decision moving with Ludwig, but he should not have expected to keep all the benefits of being a TAM fighter (which is reserved for fighters who represent the team).

Rodriguez really impressed me against Penn, but Penn is so far past his prime and hasn't even had a good round in the octagon since the Diaz/Fitch fights years ago, his best win is still probably a split decision against Alex Caceres and that was literally one fight before Penn, does anyone think Frankie would ever lose a round to Caceres?

Rodriguez is improving a lot, but Edgar fought injured against Jeremy Stephens and still won, he didn't look great against Aldo or Stephens but prior to that he was making guys like Chad Mendes and Cub Swanson look bad. I guess he's going to be worse in every fight as he gets old and Rodriguez should be better in every fight as he improves though - I feel like this fight may just be one or two fights too soon for Rodriguez, if they were fighting in 2018-9 he'd probably beat Edgar but it may be too soon for him unless he's still improving fast. I wouldn't be surprised if he hurts Frankie in the first but I don't think he can win a decision he'll need to put Frankie down for the count which nobody has been able to do yet. I just don't think I can pick against Edgar at FW until he starts falling off the cliff of old age except against Aldo (not counting Conor as a FW here since he'll never fight at FW again)

I decided to fire 3u on JDS +120 today while the price is still available as I think that fight is a flip - got bets on Maia, JDS and Joanna now and i'm planning to add Edgar, Cejudo -3.5 or UD, maybe Alvarez and then probably Branch and Skelly (or preferably Skelly decision no bet if price is similar to straight up) and maybe Aguilar too although if Carla Esparza is to be believed Aguilar was roiding pre-USADA so she may be on the downswing.

Didn't Frankie beat a similar fighter (lanky, dynamic, unpredictable) to Rodriguez minus the spinning stuff in Charles Oliveira? Oliveira has a pretty dangerous guard, however I remember vaguely Edgar having no trouble holding Oliveira down when needed and doing some GnP.
I suppose it being a 5 round fight is potentially in favor of Rodriguez since it leaves him with plentiful of opportunities to surprise Edgar, be it some magical submission or spinning stuff TKO / KO.

Oliveira is more of a grappler and Rodriguez is more of a striker is my understanding and having seen them both fight a bunch (more in Oliveira's case) it's pretty clear Rodriguez has better standup, specifically kicks and Oliveira has better BJJ/grappling

I'm pretty confident if Edgar gets on top of Rodriguez he isn't getting up, Edgar is hittable but has a great chin and great recovery and he moves well. I think I need to rewatch Rodriguez last few fights to have a better idea of how it may play out, i'm pretty sure Frankie is both the best striker and the best offensive wrestler/grappler he has faced though. Nobody Rodriguez has faced has good offensive takedowns, Penn used to many years ago but hasn't had any success wrestling/grappling since 2011 and he hasn't actually won a fight with his grappling since 2009. Frankie on the other hand hasn't been outclassed by anyone standing or on the ground in years, he did get outclassed by Aldo a bit in the rematch but it's not like it was a one sided beatdown and in both Bendo fights and Aldo 1 the standup was close. He hasn't been outgrappled since before his prime in the Maynard 1 fight which was 2008.

The only thing preventing me from going big here is how much better Rodriguez looks to be each time we see him and the fact that 35 is pretty old for a lighter weight class and Frankie should start declining soon, although it's not like he's had any awful performances, yes he lost the Aldo rematch by more than the first fight but I mean it's Jose Aldo, no shame in losing a decision there and he's looked great against everyone not named Jose Aldo since he moved to FW. It wasn't that long ago that we were talking about Conor 'ducking' him because he was the worst MU for Conor anywhere.

Does anyone have a good case as to why Branch is an underdog to Jotko? Is he going to get outpointed standing or what? Jotko's basically the least memorable contender out there, I know i've seen him fight at least 3-4 times and all I can remember is him winning close decisions against mid tier fighters. I can't recall the Leites fight at all, I sort of remember the Askham fight being a close fight. I know I watched both the McRory and Scott fights but I don't actually remember how they went down beyond what I can read on the live play by play recaps.

What I've got so far:
1.1u on Stipe @ -131
1.4u on Joanna @ -164
1.0u on Maia @ -118
1.2u on Edgar @ -142
1.0u on Pettis @ +336
1.0u on Alvarez @ +105

What I'm looking forward to when it opens at pinnacle:
- Vick, basically at whatever I can get.
- David Branch
- Jessica Aguilar
- Jared Gordon

For some reason the sherdog event page and ufc's page don't match. Did Sherman's opponent get swapped? I thought he was a late replacement to face Dmitry Poberezhets.

Wow Maia at +105 now

Single
Junior Dos Santos v Stipe Miocic
No
27/10 *
*
Will the fight go the distance?
Total Stake: (37.03 x 1)
37.03 GBP
Possible Payout (inc. stake)
137.01 GBP


Demian Maia v Jorge Masvidal
No
(Will The Fight Go The Distance - Match)
Odds:*4/7
Stake:*100.00
Possible Return:* 157.14


I am a lot more confident in the Maia / Masvidal not to go the distance, this is a typical set up I like, from previous thoughts into this, here we like two fighters extremely lethal, with DIFFERING styles,

The difference between these fighters method of winning fights is a very powerful force towards the fight ending early,

No one can touch Masvidal on the feet and no one can touch Maia on the ground,

This should be a £500 or £1000 bet here but I am working on other stuff.

Maia on the feet should not be written off, but this is also good, that he is good enough to maybe hopefully think he is good enough to stand with Masvidal.


I dislike Maia because of the bs subs he always wins by, this to me is against the sport in some way (even thou its not) I would very much like to see his face smashed to pieces but he will get his chance on the ground and then that will be all she wrote.


Santos chin is starting to go and its a deep 5 round, it went the distance last time but seeing 27/10 available with one bookie was a snap max out the line, this is a fight with similar styles and strong chins so does go the distance, or at least it did last time and I guess 6 or 7 times out of ten it does again?

lol @ i dislike maia because of the bs subs he wins by, I do lean towards it ending ITD but I could easily see Maia losing a decision standing or dominating Masvidal on the ground but not submitting him like Matt Brown almost survived 3 rounds so its possible that Masvidal could

I do think it ends ITD more often than not though just because it should be a beatdown one way or the other, it's the weird case of the pickem fight that will end up in one guy completely dominating the other most of the time just depending whether Maia can get the takedown.

I actually added a few more units on Maia at +109 and +111 to bring my price back towards evens and I also took 3u on Skelly at -125, I think he's pretty underrated and improving, I know less about Knight but his record doesn't impress me as much and i'm not sure he'll be able to hold Skelly down which is what I think he may need to do to beat him, Skelly has crazy redneck ninja flying kicks and stuff

So far I have one of my biggest ever bets on Maia, I bet across a few books but overall I have about 20 units at about -103 or something average (in bad vs current price due to smashing the -118 opener, but to be fair I still like that) and then I have 3u each on JDS +120 and Skelly -125 and 5.1u on Joanna -150. Will def be playing Edgar, I don't hate Edgar ITD either at about +5xx and will def be playing Branch too but waiting on both of those and going to keep an eye on the Cenjudo decision and maybe Vick ITD and Alvarez lines

Anyone else think it's weird that NOT Masvidal by Submission is only -2450 on 5d? I mean, is Masvidal going to submit Maia 4 times in 100 fights? Masvidal has 2 submission wins in 46 fights (granted he somehow submitted Chiesa when he was green but still impressive) and Maia has never been submitted in MMA in 30 fights and is probably the best BJJ fighter in any weight class in the UFC in terms of its application to MMA. I'd be surprised if Masvidal is more than 1% to submit Maia in an MMA fight.

I'd put the chances of Masvidal even attempting a submission at something like +1000 let alone finishing it

Added:
1.0u on Branch @ +150
1.0u on Vick @ -400
1.0u on Aguilar @ +107
1.0u on Gordon @ -160
1.0u on Benitez @ -220
1.0u on Coulter @ +133

maglame how are you modeling branch? hard to compare his recent performances with ufc caliber fights

^ right now I just consider all of MMA as one ecosystem and try to improve my fit for the whole system. Since predictions by their nature are backwards facing my hope is that improving the overall fit also improves the fit of any (reasonable) subset of the system. This assumes the overlap/crossover between groupings of fighters (by promotion, country, etc.) is large/frequent enough (it almost certainly isn't large/frequent enough that modelling several ecosystems wouldn't be better). Modelling various groupings as distinct ecosystems is definitely on my radar, but that to-do list is growing a lot faster than I can get around to things

So to answer your question directly: I make no actual distinction between fights outside the UFC and fights in the UFC. I hope that difference is captured by my performance expectations of said fighters (which I assume is, on average, lower, and my model is currently estimating as lower).

Anybody's basically at least 1% to get RNC'd from strikes or other similar nonsense in a competitively-lined fight. Unless their opponent is Kimbo. Maybe -2450 is good, but at those odds, the random nonsense submissions play a lot more than usual relative to the gameplan ones (which are likely to be 0 of any seriousness, maybe he grabs something there's no chance of finishing in a battle for position, winds up with a half-assed leg entanglement, etc)

Ended up firing 1u Vick and 2u on Aguilar today at -388 and +107, still going to play Edgar and Branch too but the steam's been against them so far so waiting for now. Going to put a unit or two on Vick ITD or SUB as well probably and maybe Cejudo decision, still deciding there.

I've decided that Maia makes or breaks my card though, will be one of my biggest MMA bets ever.

I agree regarding the random nonsense submissions, but I see fights where someone is +2000 to win by submission or whatever and think 'yeah, that's possible' which is almost the inverse of this - if there was ever a fight less likely to have a specific TKO or SUB outcome when neither fighter was a substantial favourite I can't think of one. Speaking of weird ways for fights to end, Maia himself has 3 TKO wins, one via muscle spasm, one via shoulder injury and one in his first professional fight.

Inspired by g-p's question I went and estimated/added a couple ad-hoc adjustments on top of my general model. The results were super promising! Now I'm curious if this validates my method (it works well because I have been focusing on capturing as much variation as possible on as general a level as possible), or if it invalidates it (I'm wasting my time trying to generalize too much). Now I'm kinda upset I already placed my bets for this card. This iteration might actually be close to winning

What changes if any would you make from the bets you already posted?

I ended up taking Edgar -137, Alvarez +110 and Branch +137 today as the steam against all 3 seems to have stopped. Ended up not taking Cejudo, I might if price improves otherwise I think i'm done for the card. I think the Conor fight's moved Alvarez back from being overrated to underrated again, Poirier is good but i'm not sure he can stop Alvarez's takedowns and Alvarez can hang on the feet probably at only a tiny disadvantage

^ I redid predictions and I'm on the same side on current lines, except I now find Skelly vs Knight bettable (even on current line) where I have Knight as a very small favorite. So I would have been on that at a better price (I added it for 1u).

What happened is that some of the favorites became bigger favorites and some became smaller. I guess that result is about as expected. The model is still the same, it's just fitted better.

For reference I historically pick ~73% of UFC fights correctly. I use more sophisticated criteria for evaluating my predictions, but that gives a general idea of where I'm at. That's not necessarily enough to win, obviously, but it's also not nothing. Considering where I'm at in the process I'm currently fairly optimistic.

Disclaimer: I like the idea of modelling and I wish you all the best.

I find it hard to believe that you can pick 73% of fights correctly over the long term.

As a reference point so far this year in UFC pickem I am at 56% picked correct and JimGunn is at 59%.

If you can actually pick 73% of fights correctly then you should have no problem making bank.

I am finding it tough to cap this card. I might just make a big bet on JJ and call it a day. Probably live bet some as well. These fights like Knight/Skelly and Mas/Maia are just really hard to pick.

^ You are absolutely right to be skeptical. The past is not the future. I try to judge predictions with only the information available at the time (i.e. check your predictions out of sample not just within), but for computational reasons I let some factors that appear relatively constant remain so. Then you add all the biases introduced in analyzing data etc etc.

I was just really happy when I saw that big improvement. It feels good to put a lot of work into something and get some positive result. Considering I have no sportsbetting experience, am not an expert on neither MMA nor modelling of any kind (I am close to having an economics degree, but you should see my grades. Sad!), and I have no real programming experience, I am really satisfied with getting anywhere at all.

I'm pretty sure if you disregard odds you should be able to pick 60% of fights correctly at least. That's only a -150 novig breakeven point and there are plenty of fights where one guy is -700 or whatever, auto-picking all favourites alone should yield 60%+ if you're just talking about picking a winner and not making profitable bets. This next card has a lot of competitive fights where neither fighter is -150 or bigger fav, but a lot of cards have 50%+ of the fights where one fighter is a bigger than +150 dog at which point you should be getting 60%+ on those fights and a skilled handicapper should still get a pickem fight more than 50% of the time. Actually that would be something pretty easy to determine for anyone who has a full database, what's the novig average favorite/underdog line across all UFC cards since say UFC 100?

I'm somewhat skeptical about my Skelly pick now, i'm really high on him but most of the sharp money seems to be on the other side when talking to my friends that bet. I'm pretty confident in Maia, Edga, JJ and Branch being +EV bets, less so on JDS, Skelly, Aguilar and Vick. Alvarez somewhere inbetween, if the Conor fight didn't mentally ruin him I think he's a small fav not a small dog.

Here's what I have so far, my sportsbetting roll is a lot bigger than it was when I started MMA betting but still counting $100 as a unit for purposes of the thread even though in reality my average bet size is a decent amount bigger now, I have my biggest individual MMA bet in a long time on Maia and a lot of spots I really like on this card. Down 0.15u to start due to cashing a bet out when the price improved. Decent chance i'll add either Vick by submission or ITD if I like the price when it opens otherwise I think i'm done now unless weigh ins changes anything.

25u Maia -102 (in bad because I took -118 opener; added a bunch more at +109 to +111 as well)
1u hedge on Masvidal wins in r3 +1400

6u Branch +137
5.1u Joanna -150
5u Alvarez +110
4.7u Edgar -135
3u JDS +120
2u Aguilar +107
1.6u Skelly -116
1u Vick -388

0.17u all of my picks plus Cejudo parlay +30300

Really looking forward to this card both from a betting perspective and as a fan

would expect Alvarez to bust out his full wrestling capability in this matchup. match will end ground and pound by Alvarez

Yes. % correct picks is definitely not the best measure for betting success, but it is the easiest to communicate, and is probably very highly correlated with betting success. I don't have historical betting odds in my database, but that's definitely something that would be useful.

I decided to put in the extra effort and coded a process which applies the model I'm using to data only before year x, then predicts year x, and gives me the result. I started in 2001. That means I used only data up to and including 2000, and predicted all of 2001 with that model, and so on for every year up until now. This means I'm doing perfectly out of sample predictions. Every parameter is re-estimated based only on the data available before the year which I predict. This method is fairly punishing, but should be closer to the truth. The results were as follows:

Year All UFC
2001 71% 83%
2002 71% 78%
2003 70% 80%
2004 71% 79%
2005 74% 88%
2006 72% 79%
2007 73% 73%
2008 74% 75%
2009 74% 70%
2010 71% 73%
2011 71% 71%
2012 73% 70%
2013 73% 72%
2014 73% 73%
2015 73% 68%
2016 72% 67%
2017 75% 71%

Again, probably not enough to beat the books in 2017, but its something. It is worth noting this has no special UFC adjustments. It is the exact same fight prediction model, just filtered for all vs only UFC fights.

is that picking every single fight of the year?

thats definitely beating the books since they only get low 60s% correct, like swoop accurately pointed out. sounds like youre on to something big if you can fade the variance

i've read age is a big factor in mma. are you taking that into account? also ive always wondered if some states produce better mma fighters because of HS wrestling programs