Correlation is key when picking attackers.
If everything is pointing to Conte’s Chelsea destroying poor Everton this week, don’t be cautious and play it safe by just going for Costa as your Chelsea coverage. If Costa has a good game then you can be pretty sure there will be somebody in Chelsea’s midfield that probably will too (cough* Hazard). But why stop there. If both Hazard and Costa are having good games, I expect so is Pedro, or the new summer signing Alonso. Why not include them all?
By representing 11 different teams with your 11 man line-up you are asking for 11 different outcomes you have predicted to be correct. I cant even tell you the maths on that but the % number will be incredibly low. By correlating players in teams most likely to have a strong week you suddenly spin the odds back in your favour and give yourself a better chance of winning big.
In the past two gameweeks, the combination of Firmino & Coutinho in your team would have amassed an impressive average of 94 PPG. Hazard and Costa even higher at 96 PPG. Even De Bruyne and Aguero have posted an impressive combined average of 71 points in the past two gameweeks. Correlating players likely to impact each other’s performance is a great way of improving your odds of winning.
While this method has it’s risk, if you want to win big, risk and reward is the way to go.