PyGAD v GBM

One limitation of the Sklearn library of machine learning algorithms is the inability to write your own cost function and get the algorithm from the sklearn library to us it. What do I mean by this?, well when you train a model on some horse racing data the algorithms are looking for the most effective model at predicting winners assuming you are using win/lose as your target feature ie the thing you are trying to predict. The ‘cost function’ is just a fancy term to describe the method the algorithm uses to determine how good a model is at predicting winners in our case. But what if we wanted it to do its search and model fitting based on profit or loss. What if we wanted the model fitting crtieria to be variable profit/loss ie how much a model makes betting to always win a mythical £1. The model fitting process may look at a set of parameters and find that it makes a variable stake loss of -£25 so it tweeks the model parameters and tries again and finds that the los is now -£20. Great the second model shows an improvement, lets tweek a little more.

This tweeking process will depend on the algorithm being used but as I mentioned sklearn does not permit you to define a loss function that calculates variable PL and then ask it to use it as the measure for its tweeking.

Here is where PyGAD comes in to the picture. PyGAD is a Genetic Algorithm library for Python that does allow you to create your own loss function and plug it in. An explanation of GA’s is beyond the scope of this blog entry but if you want do a Youtube search you will find some nice intro’s. There was also a great article in an old copy of SmartSig from a guy utilizing GA’s, if memory serves me it actually won the article of the year award.

OK onto the face off, how does a match up between SKlearn’s GBM algorithm and PyGAD shape up. With GBM I will be using the model based on win/lose approach whilst with PyGAD I will be adopting a model based on VPL (variable profit or loss)

For both I used 3 input features (forgive me but I wont disclose these but they are part of MySportsAI) along with Betfair SP and finishing position as data items to simply calculate PL

PyGAD

First of all I used the data from 2011 to 2017 and did not scale the input features (see later). Here are the results for the top rated horses when the model was applied to 2018 and 2019

Bets 12921 VPL -24.3 VROI% -0.79%

Next I min max scaled the 3 input features, this simply means that all 3 features are scaled to between 0 and 1 so that they are on an identical scale. This can often help with modelling for some algorithms

Bets 12921 VPL +3.19 VROI% +0.18%

My last step involved incorporating Betfair SP as an input feature. My curiosity about doing this was mainly due to the fact that when you include BFSP in a model where the algorithm is modelling for winners, the BFSP simply swamps the modelling process, its just so effective at predicting winners the algorithm will zone in on it at the expense of importance placed on your other features. With PyGAD and a cost function that is working on VPL this will not be the case. However saying that this did not appear to help as the VPL came out at -25.28 on the test data. OK worth a try.

GBM

Now onto the Gradient Boosting algorithm. GBM has appeared to be quite effective in the past on Racing data, perhaps mainly due to the fact that it can handle non linear data quite well. First non scaled input features.

Bets 14250 VPL – 12.81 VROI% -0.35%

There was obviously more joint top rated with this approach.

GBM with scaled data did very poorly, making a bad job of differentiating between rankings and delivering lots of joint or co top rated horses

bets 70804 VPL -218.7 VROI% -2.05%

Conclusion – There is room for more analysis here and encouragement that modelling on VPL could be a profitable way to go.

As always thoughts and comments welcome

Jockeys in Tight Finishes

I have just watched Joe Fanning get touched off again or at least it seems like again. Maybe I am biased and perhaps Fannings more quite finishing style keeps a horse balanced or maybe he is getting a bit old and tired its hard to be objective when you think a jockey has just cost you a few quid.

I decided to take a look at close finishes over the last 3 years, close being beaten or winning less than half a length. I do not suggest for a minute that this is a hugely scientific study but I wanted to try and find some morsel of proof to back up my gut feeling.

Here are the number of finishes, wins and percentage wins for the top 27 jockeys in terms of getting involved in tight finishes. The number 27 simply because this was a 100 tight finishes cut off point.

Jockey	Tight Fins	Wins	%wins
Oisin Murphy	232	120	51.72414
L Morris	212	90	42.45283
Tom Marquand	189	94	49.73545
David Probert	173	96	55.49133
J Fanning	168	83	49.40476
D Tudhope	167	88	52.69461
A Kirby	164	74	45.12195
James Doyle	158	75	47.46835
P J McDonald	152	70	46.05263
S De Sousa	149	73	48.99329
Hollie Doyle	145	71	48.96552
R Kingscote	144	72	50
F Norton	130	72	55.38462
Jason Watson	119	60	50.42017
Jim Crowley	118	69	58.47458
Andrea Atzeni	116	56	48.27586
Jason Hart	115	60	52.17391
R Havlin	115	55	47.82609
B A Curtis	114	52	45.61404
P Hanagan	114	60	52.63158
Rossa Ryan	110	49	44.54545
D Allan	108	53	49.07407
R L Moore	107	46	42.99065
K T O’Neill	106	51	48.11321
Rob Hornby	105	60	57.14286
G Lee	103	55	53.39806
Jack Mitchell	100	44	44

The average win percent was 49.56 so Joe is pretty average

Jim Crowley comes out top and by coincidence perhaps is a jockey I rarely complain about. A mention must be made also for Rob Hornby at 57.1%.

Perhaps rendering these numbers totally irrelevant is the poor showing of Ryan Moore unless of course you think Ryan gets beat more than he should.

Anyway apologies to Joe, you are after all average 🙂

The Derby 2021

In 1998 I sat my 3 yo son down in front of the Derby runners and asked him to pick one. He chose High Rise. I then asked his baby sitter to choose one, she was unaware of his choice and coincidently also picked High Rise. The fact that I did not put a penny on High Rise should have resulted in my reporting to the Social Services child abuse department. The whole spooky event was repeated the following year when the baby sitter picked Oath. Now of course this was nothing other than luck but I find myself drawn to the horse in the same colours as High Rise for this years Derby, Third Realm. The reasons have a little more foundation, his price of around 18/1 on the exchanges and his sectionals.

These are his sectionals for the Lingfield Derby trial with his final furlong sectional being 13sec and Sherbet Lemons, the Oaks trial winner 12.53

Now to get a handle on how these compare I divided them up into first sectional, mid sectional and final sectional which gave the following

Third Realm covered the first 4 furlongs in 49.4 compared to Lemons 54.12thats a huge 5.96 seconds difference which equates to around 0.84 of a furlong. Despite this he only ran the final third in 0.15 secs slower.

Of course that may not mean a great deal if Lemon turns out to be a Lemon but lets assume for the time being that anything winning an Oaks trial has some merit. In order to get a handle on these figures we need to dig a little deeper and so I turned to looking at other Derby/Oaks trials.

Anthony Van Dyck and Anapurna won their trials on similar ground a couple of years earlier. In actual fact I think High Realms ground was a little softer although both were classed as soft. Anapurna went on to win the Oaks and of course Van Dyck won the Derby. Here are their figures for their trials, Van Dyck is the top row

Anapurna ran to 4f out in 0.04 seconds faster than Van Dyck and yet lost 1.02 seconds on the final third. She was eased in the final furlong but never the less its hard to imagine that the figure of the final section would be reduced to around 0.15 and even if it did she is producing it off an only slightly faster first two thirds of the race. My reading of this is that Third Realm should have dipped far lower than 0.15 seconds below Sherbet Lemon and argument backed up by how the first two came away.

Finally I too a look at another pair of trials which I admit were on faster ground, Hertford Dancer (Oaks tiral) and Best Solution (Derby trial). Note here that Best Solution turned out to be a better horse than Hertford Dancer. Once again the Derby trialist is the top row.

Best Solution ran 0.5 seconds faster for the first 2/3rds of the race and paid by running 0.1 secs slower in the final third compared to an inferior animal in Hertford Dancer.

This is not an exact science I have to admit but my underlying feeling is that Third Realm did well to drop by only 0.15 seconds in the final third. The current favourite for the Derby is Bolshoi Ballet and I like him a lot, he is a worthy favourite but if the rain continues 18/1 for Third Realm looks big.

On a closer I examined the sectionals for the Dante and I was very pessimistic about the chances of anything coming from that race to the Derby.

Childs play I am sure you will agree, I am off to find a 3yo but not the equine variety

Going Going Gone

Its Saturday Lingfield Derby trial day and after a dry weather opening to the season it is pissing it down with rain and soft ground abounds. Now there is nothing that will create more column inches, amongst the failed punters we otherwise know as newspaper tipsters and columnists, than a change in the going. According to them only the four horses of the apocalypse should be feared more than a horse that has not won or run well on the ground. The most important factor from a betting angle we are told is the ground. Well I am telling you that is a load of bollocks passed down from journo to journo over the years. Its perpetuated because its easy to talk about and we can all remember a horse we waited for soft ground and when it came it won and we collected.

I am not immune to this disease having only recently paid the cost. I wanted to back Walgeist EW for the Arc de Triomphe but was put off by the very soft ground. I momentarily forgot that the ground simply meant I got a better price.

But Waldgeist and a few other anecdotes is not really evidence. Is there more robust data that refutes the going is king approach to betting. I took a look at Flat races 2017 to 2019, 3 years of data. I took all horses that had won on Turf and were racing within 30 days, so fit and in form. I then checked how they faired next time out when running on exactly the same going and when running on different going. Now if the going is paramount I would hope that those running on the same going win more often and return a smaller overall loss.

As usual I calculate returns to Betfair SP and use variable staking to win 1 point to reduce the effec of big priced winners.. Here are the results

Same Going

Strike rate 19.01% from 2530 bets

VPL -31.2 pts

VROI% -3.76%

Different Going

Strike rate 18.8% from 5720 bets

VPL +11.91 pts

VROI% 0.69%

Notice that a greater percentage of horse won from the going confirmed runners but the profit lie with the lower strike rate. This is a classic example of winners are for bragging but profit is for the bank account.

This is not of any statistic significance but here are the qualifiers for todays racing ie won LTO but now different going

1.55 Asc Lights On
3.40 Asc Keyser Soze
4.15 Asc Group One Power
4.50 Asc Pettochside
2.15 Ling Nasha Nasha
2.15 Save a Forest
2.50 Kyprios
2.50 Third Realm
3.25 Double or Bubble
2.55 Nott Mountain Brave
4.5 Bint Al Anood
5.30 Thirsk Moe Celita
6.05 Highjacked
7.35 Raadobarg

Footnote – End of day Predictions from todays blog made +2.54 pts to variable stakes to win 1 pt at BFSP minus 2% comm or for the brave +11.89 pts to level stakes

Further footnote – Although the same disparity existed in 2014 to 2016 there was no automatic variable stake profit to be had so do not think this is system to follow, the above winning day was nice from the point of view of making a point but nothing more

Federated Machine Learning for Horse Racing

The traditional approach to developing Machine Learning models for horse racing is individuals hold their data on their local machine or maybe even phone and they deploy an ML algorithm to produce a model and predictions based on that model. If we want to combine the predictions of multiple models owned by many different people we would need to either gain access to their data or access to their predictions. The former poses data sensitivity and security problems before we even consider the protective nature of individuals towards their betting data, but what if you would like to contribute to a collaboration without the need to compromise your privacy. For example what if you are a pretty good paddock judge and you would like to build a model that predicts based on horse watching (see WatchingRacehorses by Hutson). You may then want to submit the outputs of this model to a pool of models from other people which may be totally unrelated to paddock watching in the hope that your input could be a vital cog in a profitable wheel accessed only by those making ‘valuable’ model contributions. The central algorithm that takes these varied inputs would decide which are proving valuable and not only weight their input but also manage access to the predictions. It could be that your super contributor gives you a 1 minute earlier access to the final predictions than my less important predictor.

What I am describing here is an area of research known as Federated Learning and is the subject of a 2016 paper from Google. The central idea is that your data does not have to be uploaded and your algorithm is protected within your processing device. Only the predictions are contributed to the central final processing step.

This sounds exciting and enables a whole that may be more predictive than the specialised parts.

Guineas Breeding 3

In the previous two posts I looked at the pedigree implication of guineas runners over the last 10 years. In this post I have extended the data to take runners from 2000 to 2020. Using MySportsAI I have trained four Machine Learning models on the historical data (see previous posts for an explanation of the data). The ML Models are Logistic Regression, Gradient Boosting Machines, XGBoost and LogGAM. I then created the input data for this years 2,000 guineas runners and made predictions using the models trained on the historical data.

The input data for this years main runners is

trackTimeDate	horse	avgSireDist	avgDamSireDist	avgSireOr	avgDamSireOr
Newm_15:00_2021-05-01	Thunder Moon	8.65	9.89	74.23	73.28
Newm_15:00_2021-05-01	Wembley	11.11	8.04	81.89	73.67
Newm_15:00_2021-05-01	Mutasaabeq	7.33	8.99	76.38	74.72
Newm_15:00_2021-05-01	One Ruler	9.19	7.76	82.82	76.02
Newm_15:00_2021-05-01	Master Of The Seas	9.19	8.31	82.82	72.66
Newm_15:00_2021-05-01	Battleground	8.35	9.6	84.04	76.68
Newm_15:00_2021-05-01	Poetic Flare	8.78	8.44	71.55	71.27
Newm_15:00_2021-05-01	Lucky Vega	8.74	8.27	76.16	73.65
Newm_15:00_2021-05-01	Chindit	8.34	7.04	70.73	67.93
Newm_15:00_2021-05-01	Mac Swiney	9.78	8.33	80.75	69.47

The following is the results after deploying the four ML algorithms with sum total of rank positions given by each algorithm. Overall the lowest total ranked position would be the top rated across the four models.

TrackTime	Horse	LogRegRank	GBM	LogGAM	XGBoost	Total
Newm_15:00_2021-05-01	Battleground	1	1	5	1	8
Newm_15:00_2021-05-01	Thunder Moon	3	4	6	3	16
Newm_15:00_2021-05-01	Mutasaabeq	7	3	4	2	16
Newm_15:00_2021-05-01	Wembley	2	6	1	8	17
Newm_15:00_2021-05-01	Mac Swiney	6	7	2	5	20
Newm_15:00_2021-05-01	Lucky Vega	8	2	8	6	24
Newm_15:00_2021-05-01	Master Of The Seas	4	8	3	10	25
Newm_15:00_2021-05-01	Poetic Flare	9	5	9	4	27
Newm_15:00_2021-05-01	One Ruler	5	9	7	9	30
Newm_15:00_2021-05-01	Chindit	10	10	10	7	37

Top rated therefore is BattleGround followed by equal 2nd Thunder Moon and Mutasaabeq

I will run off the 1,000 guineas predictions when the decs are known, watch this space

You can find out more about MySportsAI at http://www.smartersig.com/mysportsai.php

Footnote – Insufficient data to access Van Gogh

Guineas Breeding 2

In the previous post I looked at how the sires and damsires may predict the ability of Guineas runners to handle the 8f of the Newmarket track. The message seemed a bit muddy but some possible trends may have been evident. In this post I will turn the attention to sire and damsires predicting class as measured by official ratings. The same approach will be used as the previous blog post but here the OR where recorded will be used as the measure and of course averaged when looking at Guineas runners.

OK lets get onto the data and results, first of the sires, the average place OR of sire offspring for all guineas runners was 75.86. Using this as initial split ie those runners where sireOR > 75.86 and those <= 75.86 we have the following.

	Wins	Runs	%	Vpl
>75.8	12	107	11.21495	2.741398
<= 75.8	4	100	4	-0.40339

Quite a clear split between the two and generally as you increase the OR threshold the win percentage gets higher

OR threshold and percentage wins

Onto the dam sire data

	Wins	Runs	%	Vpl
> 72.2	8	110	7.272727	0.677137
<= 72.2	8	97	8.247423	1.660867

The dam sire influence is less pronounced or should I say not significant at all.

Closer to Guineas day I will run off the various breeding stats for this years runners and finish with a final Guineas blog. After that onto the Derby to see if one can profit from pedigree.

Guineas Breeding

Tags

2000 guineas breeding, Horse racing breeding

This morning I have been browsing over the stamina influence of various 2000 and 1000 guineas runners since 2010. Its a topic that will gather momentum as we enter the classic period and none more so than the Derby which I may blog about in a week or two if there is interest in this entry.

My approach was to analyse the placed percentages of all sires and damsires by race distance up to the point of the guineas and then see if there was any influence. For example when rounded down to the nearest whole number did runners with 8f sires do better than those with 6f sires, also what role did the dam sire play if any.

The results are somewhat confusing but interesting, as with all data if you look hard enough you may find something and its possible that there may be the odd clue hidden away. As in all betting data we are also faced with the question, are we looking for winners or are we looking for value. What if the 8f sired horse win more often but the public seem to know and as a result the wait for a 6f winner can be long but more than pays for the losses. I have calculated PL to Betfair SP after 2% commssion and the PL is variable stakes ie to win £1 whatever the odds.

First the sire performances

Sire dist	runs	wins	%	AvgBFSP	Vpl
6	5	1	20	46.424	0.836727
7	45	3	6.666667	116.8876	0.571075
8	28	0	0	77.24893	-1.97151
9	32	3	9.375	31.62375	0.804533
10	15	0	0	133.8687	-0.65128
11	23	6	26.08696	15.86087	-0.65128

To my surprise grouping the results from 6 to 7f and from 9 to 11f resulted in a better strike rate for the latter but the profit was in the 6f and 7f sires

Now on to the dam sires

Damsire	runs	wins	%	AbgBFSP	Vpl
6	5	0	0	188.33	-0.18057
7	15	0	0	103.63	-0.8818
8	65	8	12.30769	63.51431	2.768976
9	42	2	4.761905	67.75548	-1.4114
10	18	3	16.66667	78.65389	2.360995
11	2	0	0	73.285	-0.04751

With the damsires during the last 11 years you really needed to be on the 8f to 10f dam sires

Closer to the day I could produce the data for this years runners if you are interested. In the meantime I would love to hear from those with a keener interest in breeding than me and perhaps get their insight into the subject. Good luck if you are having a bet on either or both the guineas

The First Rule of Betting

Stumbled on a tweet the other day asking

“What is your best advice to a first-time bettor?”

There was a lot of good responses such as manage your bankroll, bet on your most strongest opinions, do not chase losses and so on some were far more esoteric. Nothing wrong with these suggestions but I wondered, are they really the lessons that will get you to first base?. As for back your strongest opinions well if you think a novice punter can figure that out then please explain in the comment section below.

My response was that the first fundamental mind shift that a punter has to make in order to get them to first base. Something that will at least mean they are are on firm ground when looking to get to second base and one paradigm shift that hopefully with some examples and suggestions can be made by everyone, is to understand that its all about the price.

What do I mean by this?, well its the basic idea that unless you are backing selections that are over priced ie bigger than their true chance of winning then you are not going to make money in the long run. Its as simple as that, really. Just about everybody will agree with that but the vast majority have trouble implementing this approach. To bet on something that you think is too big when you do not think its the most likely winner is a difficult thing get your head around and action. In fact its worse than that as the vast majority do not even think in terms of chance they stick to the simple idea of what is going to win and then they decide to bet on it if the payoff would titillate them or not. At even money it hardly seems worth it but at 3/1 I could buy a new suit.

Value the word I have avoided using, a word that even some pundits on TV will fail to understand. How does one evaluate value, how do you arrive at value bets. There are two main approaches.

The first approach is the more nuanced and requires a certain skill set. It involves calculating the chance of say each horse in a race and then comparing that chance to the chance the exchanges are offering and bet or no bet depending on whether you think the selection has a better chance than the odds indicate. Calculating these chances or probabilities can be done via a couple of methods. The first is seat of the pants experience based on form reading and knowledge of individual horses. This takes time and experience and puts a lot of emphasis on your own strength of mindset after all when things are going poorly you somehow have to operate in the same consistent manner. This is very difficult do but there are people who can simply spot over price selections. The disadvantage to this method is that you probably generate fewer bets which can be a problem for reasons outside the scope of this blog post.

The second method within this first method is to create statistical models based on input data in say horse racing if that is your field of betting. This model can then generate probabilities for each horse in a race. The benefit of this approach is that it can be automated and can produce more turnover in bets.

The second method which can produce value is to not attempt to calculate individual odds on horses but to try and find categories of horses that are overpriced. For example you may find that Richard Fahey’s horse when ridden by a claimer are over priced and long term return a profit. You do not know which ones are over priced you just know they are as a collective. Of course this can change as can the relative weights or merit of input values to a model. Also with a model you do not know which horses probabilities were correct and which were wrong you just know they work as a collective. The key difference thought is that with the Fahey runners you are backing them regardless of price whereas with a model you will skip some horses because of the price.

Now I am not suggesting that the last few paragraphs should be thrown at a novice bettor but its abolutely fundamental to betting that a new punter gets their head around the value concept. Pretty much all else is lost if they do not grasp this and base their betting upon it

If you found any value in this blog post let me know via the rating below and comment section

Pre Cheltenham Going

With Cheltenham only a couple of weeks away some concern was expressed on the MySportsAI email forum about the extreme prevalence of soft and heavy going this NH season and how this may cause upsets at Cheltenham when we are likely to hit better going. Could all those fancied horses who have done well over the winter on soft going come unstuck at Cheltenham when better ground is likely. If this is the case then bookmakers could be in for a good week.

To examine this I looked back at the last 10 seasons and did a simple calculation. Using a numeric value for the going where Heavy = 6 and Firm = 1 I calculated an average going for the months November to February in each year. I then examined the fate of the market in that years Cheltenham festival. Perhaps higher average pre Cheltenham going would mean more market upsets.

Here are the results for all Betfair SP prices and laying to win £1

Year 2010 pre going 4.62 LayVarPL +2.174 average win price 20.84

Year 2011 pre going 4.43 LayVarPL -0.61 average win price 15.2

Year 2012 pre going 4.14 LayVarPL -0.02 average win price 14.46

Year 2013 pre going 5.03 LayVarPL -1.54 average win price 15.98

Year 2014 pre going 4.8 LayVarPL +0.06 average win price 17.27

Year 2015 pre going 4.78 LayVarPL -0.42 average win price 13

Year 2016 pre going 5.01 LayVarPL +2.17 average win price 10.7

Year 2017 pre going 3.05 LayVarPL -0.61 average win price 17.1

Year 2018 pre going 4.56 LayVarPL -0.12 average win price 15.12

Year 2019 pre going 3.68 LayVarPL -0.35 average win price 24.99

As we enter 2021 Cheltenham the pre going value is 4.54 not as generally soft as some other years.

The above figures do not seem to spell a pending doom for punters at the front of the market, indeed the best year was 2016 which came off a seemingly wet 5.01 winter. As I have always said, in general terms does the going really matter, well not really.