We're sort of winning! Stats and projections before Arsenal-Wigan

This weekend, Tottenham Hotspur and Chelsea both escaped tough road matches with three points, both on winning goals past the 80 minute mark, both coming with the club on the front foot following a sending off earlier in the period. Chelsea's win put them through to the Champions League barring some crazy 10-goal games next Sunday, while Tottenham's win just put the pressure on Arsenal ahead of their Tuesday clash with Wigan Athletic. Arsenal have fallen behind Spurs in the table, and they need a victory to re-take fourth place. We're sort of winning.

One could argue that Wigan will be less motivated after their FA Cup victory on Saturday, but the results at the bottom of the table Sunday should provide more than enough motivation. Newcastle and Norwich City both ensured their safety with huge wins. Southampton's draw locks up a place in the Premiership next year, barring nutso goal scoring on Sunday. Fulham lost, but Aston Villa's loss on Saturday had already clinched safety for the Cottagers (again barring goal difference insanity). The result is that only three clubs are realistically at risk of relegation—Wigan, Sunderland, and Aston Villa—and of those Wigan are by far the favorites to go down.

I can make it simpler. Wigan must win both of their next two matches to survive. They have to win at Arsenal and then beat Aston Villa on the last day. This makes for one interesting twist for Spurs fans. A Wigan victory is basically no different from a Wigan draw for us, but for Wigan it's the difference between possible safety and near-certain relegation. They will not sit back and protect a draw if it gets to that point, as much as we might like them to.

Projected Table

Aston Villa actually project fractionally ahead of Fulham on points, but Fulham are almost certainly safe because of their superior goals scored / goals conceded numbers. If Wigan win out and Sunderland get at least a point from Spurs, then Villa indeed will be relegated on goal difference. In that case, I can pretty much guarantee that we'd feel bad if Aston Villa were relegated on goal difference.

So, Arsenal-Wigan is going to be a huge game for both the bottom and top of the table. Here are the likely outcomes and implications of the game.

Projections for Arsenal-Wigan

I already ran the match projection table for this one in my last update—the quick version is that an Arsenal win is by far the most likely outcome at 69%. 16% draw, 15% loss. Here are the effects of those different outcomes on the top four and relegation races. I list under the W/D/L headings each club's chance of making top four and chances of staying up in the Premier League based on those outcomes for Arsenal on Tuesday.

There actually was one simulation (#2119) in which Wigan drew with Arsenal on Tuesday and then Sunday went crazy. Wigan won 6-1 over Aston Villa, and Spurs beat Sunderland 7-1, causing Sunderland to be relegated on the second tie-breaker (goals scored). But otherwise, it was win or go home for Wigan.

Obviously Tottenham are big favorites to beat Sunderland. Some folks worried that because Sunderland's draw on Sunday left them at risk of relegation, the Black Cats may be desperate for points in their match at the Lane on the last day of the season. This is true. However, the only way for Sunderland to have anything to play for on Sunday is for Wigan to beat Arsenal mid-week. If Wigan draw or lose, Sunderland will be safe regardless. So the only scenario in which Spurs play a desperate-for-survival Sunderland side will be a scenario in which we head into the final day of the season in fourth place, in control of our own destiny. So it's sort of win-win. Unless we lose.

The Stats, Juked

On Friday, I put together a second set of projections based not on a club's play over the full season, but only since January. I did this in order to make the numbers look better for Spurs, since Wigan and Villa have played better since the new year, while Stoke have been a lot worse. This method still ranked Chelsea, Spurs, and Arsenal wins as the most likely outcomes, but it increased the chances of good outcomes for us and bad outcomes for them. Chelsea's win means that it doesn't matter any more how Aston Villa project, but the Wigan Athletic side that beat Manchester City in the FA Cup final sure looked like at least an average Premier League side. Hopefully they can play that well again.

If Wigan project as a reasonably good club, as they do based on their numbers since January, then the chance of Arsenal winning on Tuesday drops to 60%, and the chances of a draw or loss increase to 19% and 21% respectively. This is how the projected table looks after that minor juking.

This one doesn't make such a big difference because the Stoke and Aston Villa games are now out of the way. But it's a few percentage points. Good enough for me.