Monday 21 October 2013

Results Optamisation

The debate about goal expectancy based pricing models rumbles on, with much of the conversation having moved to that of in-play models. To reiterate my opinion on trading football in-play, unless you are an expert reader of football and / or have a sophisticated model that can price the game accurately as events unfold, your time is better spent elsewhere. The average person is highly unlikely to be able to consistently find an edge. There is a lot of smart competition out there. Pre-game, the playground is much fairer, and as I wrote before, even though you can never predict how any single game will unfold, you can predict how they will turn out 'on average'. Anyway, here are some comments, starting with James who wrote:

Your goal expectancy should already include the 'game state' information - just not with the correct weightings.
I would like to know where Emp gets the data for all of those highly different situations. As far as I am aware the only place to get goals data (and more importantly shots data) given the game state are companies like Opta. Unless anybody else knows any good sources?
I'm not sure that there would be much improvement anyway, or should I say enough to justify the extra effort required.
James then added:
On a separate note - does anybody combine the XX draws selections with any bookmaker promotions?
I have no idea how people use the XX Draw selections. I imagine the profits from them are used for any number of quality of life enhancements, but how and where they are applied will likely vary. For some, they are probably best used to mitigate the Betfair Premium Charges, and some may well use them as part of a portfolio approach. Comments welcome, preferably after we have had a profitable weekend!

As for the first part, I think the Pareto Principle comes in here. As James says, it is likely that very little improvement would come from the significant effort needed to handle all of Emp's game states. Several full-time enterprises are engaged in this, and most of us do not have either the time or resources to get close to competing.

Anonymous says:
Think you may have missed the point Cassini.
Expectation of a goal is known to be dependent on the current score and there are variables in a game that can not be controlled that will accelerate and impede goal production.
It is the game state that will effect accuracy and shot on target production so if for example the game is 1-0 at half time to the home team you can look historically to see how this effects shot on target production to the home team and the away team in the second half and again should the game go 1-1 at any time you can repeat the process.
No model is perfect but the trap you are falling into is to not accept that in play models are far more precise then the static pre off models because you can react to the events such as a red card and an early away goal and 1-0 half time game state if you understand how they effect shot on target production and accuracy.
I never meant to imply that pre-game modes were more accurate than in-play models - only that as said earlier, the effort to develop the former is a lot less than the latter, and that the in-play model is up against serious competition.

Emp clarified his analogy of cricket and run expectancy by saying:
Sorry for the added comment, but I feel that I may not have made my point in the best way when giving my cricket analogy.
What I was trying to say with cricket is that playing styles depend heavily on situations. For instance, you can compute M.S. Dhoni's strike rate and averages, and contrast those with the bowlers economy rate and average. That particular calculation won't yield a reliable prediction of whether a team of 11 M.S. Dhoni's is likely to achieve a particular target. Why not? Because his averages and strike-rates are a cumulative of his performance over various highly different situations. Batsmen don't bat the same way with 25 overs and 5 wickets left as they would with 5 overs and 7 wickets left.
Similarly, and this is the point I was making, football teams don't play the same way when they are leading by two goals and when they are level with 15 minutes left. Stronger teams in particular, are disproportionately likely to score goals (at a rate above their goal expectancy) when the situation absolutely requires them to score.

There's no question at all that a highly accurate goal expectancy model (like Cassini's) will be very profitable, there's no way I was suggesting he should scrap it. What I was saying that an equally sophisticated model that measures relative strength of teams by analysing results (not based on these goal expectancy methods) could be equally effective.
I still contend that any model which ignores goal expectancy is, by definition, unsophisticated. Results in football are very often inaccurate reflections of the game and analysis of results may be of academic interest, but it is not a solid footing for making investment decisions.

Anonymous had this to say about Emp's comment:
"Similarly, and this is the point I was making, football teams don't play the same way when they are leading by two goals and when they are level with 15 minutes left. Stronger teams in particular, are disproportionately likely to score goals (at a rate above their goal expectancy) when the situation absolutely requires them to score."
There is no data to confirm that the above is true .
"What I was saying that an equally sophisticated model that measures relative strength of teams by analysing results (not based on these goal expectancy methods) could be equally effective."
I worked for a football trading house and for pre off they use POWER Models i.e. player rating and in running they use Shot Strength data which is manually keyed .
All models are based on relative strength of teams or goal expectancy.
What other method would there be?
What Emp confirms is the lack of knowledge of the effect of game state during the game for the simple reason that people do not look at the data.
There is data and there is interpretation.
Unusually, I fully agree with Anonymous on this, - "all models are based on relative strength of teams or goal expectancy".

His "football trading house" comment also confirms what I have said about the competition that you are up against in-running. That's not to say pre-game is easy - just easier, in that you have more time to analyse data, and that prices are far less volatile than in-play prices.

No comments: