SHL stats – xG table, leaderboards and top scorers

How to read the model’s table

Start with xGD. A team with a positive xG difference creates more than it allows, and sooner or later that usually shows in the points column. A team high in the table but low in xGD is living on the margins: a hot goaltender, shots in off the crossbar, a power play converting 30 per cent. That kind of thing regresses. Not always next week, but over a season.

That is where the value is. The odds market still prices a lot on league position and form in results, not on the underlying numbers. A SHL team that has lost three in a row with a positive xGD is cheaper than it should be. That is the simple idea behind everything we do here.

Shots on goal and save percentage go together. A team that allows 32 shots per match but has a 93 per cent save rate looks good in goals conceded. Until the day the goaltender is human again. Check both columns, not just the one that looks good.

Playing style is not a grade. Shot volume does not win more often than puck possession. But it tells you something about how the team wins – and which markets become interesting. A shot-volume team drives shot totals, a transition team scores more in the third period when the opponents are chasing.

Leaderboards that measure more than points

The goals and points leaderboards are here as they always have been. Next to them are the xG and xT leaderboards, which are more interesting. A player with six goals on 2.1 xG has finished brilliantly or been lucky – usually a bit of both, and it rarely lasts. A player with high xT and no points creates the danger but does not get the reward. Yet. That kind of player is often mispriced in the goalscorer markets.

The per-game-unit column normalises for ice time. A fourth-line centre playing twelve minutes a match cannot be compared with a first-line forward on twenty – not in totals. Per game unit he can. That is how scouts read statistics, and that is how we read them.

The goaltenders are ranked on GSAx – goals saved above expected. Save percentage punishes the goaltender behind a poor defence. GSAx looks at the shots the goaltender actually faced and asks how many an average goaltender would have let in. The difference is the goaltender’s credit. Or blame.

From stats to predictions

The numbers here are the same ones the model uses before every match. The difference is that the match page combines them with home advantage, form and opposition into a probability – and compares it with the odds. Go to this round's predictions, pick a match in the SHL fixtures or see how the predictions have performed so far.