EPL 19-20 September 2015
EPL 19th-20th September 2015
Another weekend in the EPL, let's look at the basic statistics for all the games:
%matplotlib inline
import league_analysis
from IPython.display import display, HTML
epl = league_analysis.epl
league_analysis.display_matches(epl, '19/09/15', '20/09/15')
Chelsea 2-0 Arsenal
The red cards, and in particular Gabriel's just before half-time, somewhat robbed us of an otherwise entertaining and interesting football match. Subsequently Costa has been given a ban for 3 matches whilst Gabriel's red card appeal was successful.
So there is certainly much debate to be had as to whether Chelsea really deserved their victory. However, this match represents the first match of the new league season in which Chelsea have really (shot) dominated their opponents. Likely a large portion of that is due to the red card just before half-time, but even before that Chelsea looked the better side (though we do not have the data here to back that statement up).
In any case, dominating shots 22-9 and shots on target 7-2 is a performance worthy of a victory, whether or not that performance was achieved through dismissals and/or gamemanship. This now makes the rest of Chelsea's season intriguing.
So too for Arsenal, who have generally dominated teams, by a significant margin. They have had some trouble converting shots into goals, but not performing against Chelsea is likely a more acute headache for the club.
Bournemouth and Tottenham
For the second week in succession Tottenham earn a well-deserved, if nervy, 3 points. Again they have dominated the shots, but again failed to really put the game to bed, winning only 1-0. Still, Tottenham seem okay at the very least, and are arguably so far yet to reap the rewards their performances really deserve.
Bournemouth are a great team to watch and were one of only three teams this weekend to unambiguously (at least in terms of shot numbers) deserve their victory.
Alarm bells have been ringing all season for Sunderland fans, but last weekend, despite the loss to Tottenham there was a general air of optimism that the worst may be behind them. Not on the basis of this performance against the league newcomers. So far, Sunderland, firmly in relegation form.
Man City 1-2 West Ham
Manchester City continue to dominate every team they have come up against. This time though, they did two things they have not done all season. Firstly, concede a goal and secondly fail to win. West Ham really are riding some incredible luck at the moment. I still don't expect that to continue.
Liverpool 1-1 Norwich
Few people are making that case that Liverpool look like a good side capable of challenging for the top four. A draw at home to Norwich won't convince anyone to start. But the shots stats at least suggest they were dominate and should really have scored a victory here.
Aston Villa 0-1 West Brom
Never nice to lose a derby at home. We see that the game was at least pretty even, but West Brom deserved to win more than Villa, a worrying sign for the home side. Villa, like Newcastle and Sunderland, will be very worried at their start, especially since the three promoted teams all seem to be handling the top tier if not comfortably at least with optimistic signs. Add to that, Leicester's good form despite having propped up the table for much of last season and you're starting run out of relegation candidates.
Newcastle and Southampton
Both lost home games. Both had the lion's share of shots and shots on target. Newcastle only marginally and they were behind for a large portion of the game. Southampton came up against Manchester United who seem to be defying much of the statistics this year. This is the second successive game Man United have scored 3 goals from 3 shots on target. Again Manchester United really seem to be doing something to defy the stats (but I'm open to being wrong about that). On the other hand Newcastle need to turn something around and quick.
Stoke, Leicester, Swansea, and Everton
Lastly two draws which were more or less from even games. Any of the four sides could have nicked a victory. I think Leicester will be the happiest of the four given that they were 2 goals down.
Another point there were a few games in which the team with the higher number of total shots had the lower number of shots on target. This usually means a pretty close game, as in the case with Swansea and Everton. Aston Villa will be more worried that they managed 13 shots only 3 of which were on target, whilst allowing 6 shots on target from their local rivals.
Changing Statistics
We've reached game six in the premier league, so it is now time that we can start to look at some trends in statistics. Mostly for demonstration purposes, this is a graph of how a team's points have increased over the first 6 games. You'll see each line in this graph always increases, or is flat (if the team in question lost that particular game).
after_game_no_dicts = league_analysis.collect_after_game_dicts(epl, '01/08/2015', '20/09/2015')
league_analysis.plot_changing_stats(epl, after_game_no_dicts, 'points')
Total Shots Ratio
So now we can plot how the total shots ratio has changed over the course of the first 6 games. Of course this is partly a plotting of the teams each team has played. But still you can see some trends emerging. A reminder that the total shots ratio is simple the proportion of shots seen in a team's game that they themselves took. In other words if you have a total shots ratio of 0.5 then you have taken the same number of shots as your opponents have taken against you. If you have a total shots ratio of 0.75 you are taking 3 times as many shots as your opponents.
league_analysis.plot_changing_stats(epl, after_game_no_dicts, 'tsr')
So in particular we see that Newcastle and Sunderland really are pretty abject, at least as far as taking/preventing shots is concerned. This suggest their league positions are relatively true to their form (at least relative to their opponents, it could be that they have faced particularly stiff competition).
Ranking Stats
We can additionally rank the teams in terms of a particular statistic. This can sometimes be a little easier to see that looking at the final positions on the graph, but displays the same information. In this case we display the TSR for teams and also show their points.
league_analysis.display_stats_table(after_game_no_dicts, ['tsr', 'points'])
Perhaps the most interesting point here is that West Ham are truly over-scoring for their performance. They are 18th in the league for TSR but have 12 points for 3rd place in the league table. Crystal Palace as well might consider themselves a little fortunate. Leicester and Manchester United also seem to be doing well for points compared to their TSRs.
On the flip side, Arsenal, Southampton and Tottenham all seem like they could have achieved more points but for a bit of fortune. Manchester City are fine and deservedly top the table, Chelsea doing about as well as they should have.
Comments
Comments powered by Disqus