So having just posted something backing up a uMAXit piece I wrote, Simon Gleave made a suggestion about incorporating shot conversion rate against. So, I have done that. I’m going to call it Shots on Target Coefficient – STC = (SOT/SOTA)*(SCR/SCRA)
In other words, STC equals shots on target divided by shots on target against, all multiplied by shot conversion rate divided by shot conversion rate against. I have calculated SCRA by dividing goals conceded by shots on target faced and then getting a percentage from that.
For the 2016/17 season, as it currently stands (after 23 games played):
|West Bromwich Albion||74||109||13.19||29||26.61||0.337||8|
|West Ham United||78||115||9.29||40||34.78||0.181||11|
The value of R is -0.8742. This is a strong negative correlation, which means that high X variable scores go with low Y variable scores (and vice versa) – this is obviously because a high number league position is a bad thing, but you figured that out.
Interestingly, factoring in Simon’s suggestion, the correlation is stronger than without it, -0.8742 compared with -0.8417 – factoring in SCRA makes this model even better.
I’ve not yet run this for the other seasons, but once I do, I think it’ll work.
This model, interestingly, suggests that Bournemouth are overachieving, Palace are underachieving (Simon suggests due to their propensity to concede first), and West Ham are massively overachieving, a real worry for them now that their main creative source Payet has left. Burnley are overachieving because of their amazing defensive solidity, especially Tom Heaton – while they might regress, they should be safe. Sunderland are doomed. Surely. Right?
Thanks to Simon Gleave for his suggestion.