Using the stats

[cde]

I was thinking it might be helpful to hear how people use the player statistics to understand and shape their teams and players.

Picking players based on good averages if fairly obvious but by using some of the other information you can better choose who to put through training, how pairing certain bowlers or just a change in the batting line up can make a big change to success.

I've had a few lightbulb moments over the years I've been playing Cricket Captain and would be interested to hear other people's.

I would have to add the caveats that in the game, as in real life, there are no sure things just ways to improve your chances and that the way player improvements work through the coaching system can be fairly random.

[cde]

I statrted thinking about it after reading a post the other day along the well known line of "What does the AI know about a player that I don't?"

I can't quite remember where it was but the upshot is that a few seasons in they had a player they clearly (quite rightly) liked who regularly hit 100s and was a star batsman but the AI kept calling up batsman B who had about the same average as A but hit far fewer 100s. What was it that made the AI keep choosing the batsman who scored few 100s over the other batsman?

The AI does seem to make some bizarre selections but this one was probably about right and a couple of years ago after I finally twigged what is happening in these scenarios and how to make it work for me.

To illustrate what I mean I will use round numbers and exagerate a bit, but it isn't complicated when you recognise it on screen.

In 10 innings batsman A averages 50 with two 50s and three 100s. B averages 50 too but with four 50s and one 100.

So both reach 500 runs but batsman A scores at least 80% of them in just 5 innings meaning that he averages, not allowing for ducks, just 20 in those 5 innings. Batsman B scores at least 60% of his runs in those 5 innings meaning that his average in the remaining matches is 40.

So really you have two very different batsmen, one who is good at converting 50s to 100s but is more likely to get out cheaply and one who will reach about the 50 mark almost every inning quite reliably. About half the time this is down to one being an aggressive and the other a less aggressive player, where one is more likely to swipe at a ball best left alone.

But this is not always the case and often they are of similar aggression - and this is why you notice one getting picked over the other. I used to think it obvious which is the better out of two similar players and blame the AI.

What I found interesting though is what happens when you pay more attention to both batsmen during matches and try to improve them through training.

Aggressive batsmen do tend to go big or get out cheaply but that is not always the case here, especially if they have a roughly equivilent strike rate. After paying him more attention you often see that B is a little less likely to get out cheaply but it is more noticeable that he keeps getting out for between 70-90. So rather being Mr. Reliable and scoring about 50 each time he is actually failing to convert 50s to 100s.

This can be remedied and he can become every bit as good as batsman A at scoring centuries by giving him general batting training. Like all the training this can take a while, or not happen at all but I've found that you can increase the chances of it happening a little by waiting for the right moment. If they fall short of a century a couple of times in a row then start training then. It absolutely does not work every time but sometimes the improvement can be quick and they suddenly get over the line and you find that their average picks up as they get more 100s.

The first time I caught on I realised that the AI was right, this was the better player but I wasn't spotting what the statistics were telling me. I'm sure there are lots of clues in the player ststistics that I am missing and some of the things I put down to odd AI decissions are actually me not knowing what to look for.