Psychology isn’t something that is limited to the minds of traders, it’s out there on the sports field and in the associated markets as well. I’ve observed all sorts of strange biases that just don’t have any place in a rational world, but feature heavily and persistently in the markets and in the field of play.<\/p>\n
The most remarkable thing is that these biases just seem to persist and persist and there seems to be little that people can do about it. It’s just human nature.<\/p>\n
Psychology in Tennis<\/strong><\/p>\n Tennis more or less boils down to how people play key moments in the match. When I realised this I thought I would start examining those keys points to understand what happened around there. As a precursor to this, I look at just individual games.<\/p>\n The question I was trying to ask was does a player play 15-0 the same they play 0-15 and what happens when they reach break point? The answer, as I often find, was never straightforward.<\/p>\n Thought experiment<\/strong><\/p>\n Before I analyse anything I try not to think about the outcome as you can accidentally influence your data mining. But you have to work out how you will structure your data and what you expect to find.<\/p>\n I figured that a professional Tennis player would play each point to the best of their ability. 40-30 shouldn’t be that different from 40-15 and when going for the game, maybe, in fact, they would step it up a gear to close out the game.<\/p>\n Before reading the rest of this blog, what do you think? Do you think all points are equal or are some point more equal than others?<\/p>\n Before I analyse anything I try not to think about the outcome, as you can accidentally influence your data mining. But you have to work out roughly what you are trying to do. I considered looking the odds at those points in the match but I can model what should happen well enough so I needed to find some other measure of what was going on the in the mind of the player. I settled on service speed.<\/p>\n Using average service speed would give me a clear measure of how a point was being played. If the server was confident they would serve more aggressively and faster. If they were nervous and worried about getting a shot in, they would serve much slower.<\/p>\n The results<\/strong><\/p>\n I carefully picked through reams of data and I initially focused on 40-0 and 0-40 as that seemed an obvious extreme. Sure enough, the average service speed was different but in the ‘wrong’ way. I had imagined that maybe at 0-40 they had nothing to lose by trying to get a decent serve in to get back into the game. But it turned out that it was the other way around.<\/p>\n To confirm this wasn’t an anomaly I went through each of the score lines and ranked them according to how many shots it would take to finish the game and there was a correlation. The further ahead a player was they seem to serve harder. The further behind they were they would put in a weaker serve.<\/p>\n It sort of makes sense that they would want to get the serve in, but by making it weaker they would give the returner a better chance of finishing off the game.\u00a040-0 recorded the fastest average serve and 0-40 the slowest. The deeper trouble the player is in, the slower the service. That sort of makes sense, but doesn’t seem logical in a game plan. Surely you should be risking it a little to get out of trouble?<\/p>\n It varies a bit up and down the game but it’s an interesting thing to observe. I’ve reproduced a summary below, I’d be interested in your thoughts if you want to comment. The interesting thing for me was that I almost expected it, but didn’t. A professional player shouldn’t vary the shots that much if they are playing at the highest level, but they do. Which implies that even though they look cool and calm there is some torment going on underneath the surface.<\/p>\n![\"\"](\"https:\/\/www.betangel.com\/blog_wp\/wp-content\/uploads\/2017\/08\/2017-08-29_09-20-31.png\")
<\/p>\n