***WARNING - THIS THREAD IS ABOUT DATA. FEEL FREE TO IGNORE IT IF YOU THINK DATA MEANS NOTHING*** I found this interesting so thought you might find it to. Would be interested in the views of others to see if I'm interpreting it right. Particularly @SourPatch2 as he seems to understand data better than me This graph below compares the speed teams move up the pitch on the Y axis. And the volume of passes in their attacking moves on the X axis. Anybody top left of the chart is effectively a long ball merchant. Anybody bottom right is looking to dominate possession and move it about their back line. Sunderland's position Weirdly we find ourselves at the bottom left of the graph. Meaning we have little possession but we are also the slowest team in the league at moving the ball up the pitch when we do have possession. (obviously apart from the Mundle goal last weekend) My theory was that this is because we are the best team at winning possession up the pitch. We don't need to move the ball fast up the pitch, because we're winning it in attacking areas anyway. Often we'll actually win the ball and go backwards before we go forwards again. However Opta data actually shows we are one of the worst teams at winning possession high up the pitch. We've only done it 17 times versus Stoke's 34. (see attached screenshot) So what on earth is going on? please log in to view this image
Just from memory, I think we win a lot of our turnovers in the middle of the pitch after forcing a bad pass into the midfield by pressing high, the Mayenda goal against Sheff Wed for example where we forced the pass into Windass and produced the turnover on the halfway line. Think gamestate plays a large role as well, like the Burnley game for example we pressed really high for 30 minutes, scored and then settled into our shape for the rest of the game. Whereas the teams that have high pressing numbers are often chasing the game.
Cos "data" does not play football or watch it with a pair of eyes. Only data that is certain is we are top of the league
I can remember when football was a simple game…when you’ve got the ball put into their net….when you haven’t got the ball stop them from doing that to you….simples really.
Yeh guys - this thread is about data. If you don't think its important you don't need to come into the thread.
The only thing data misses sometimes is ‘intent’ . Won the ball less than Stoke, but that’s because we have attempted less. Low on fast breaks? Cos it’s patience and doing so when the prime opportunity has presented itself. Data is great, for half the story. When we have broken fast, we have exploited space and had a man over. Other teams might have broken more times but had no free man or out ball. Then, the oppo is expecting it next time and sits off. ‘smarter not harder’. You have to combine the data with the real world scenarios for the reality. Also, have we only scored once from pressing? Clarke v Cardiff, Mayenda v Sheff Wed…
Love this view. Look how wide our central midfielders are. Creating that overload / overlap we saw RLB encouraging in training videos during pre-season please log in to view this image
That looks much more aligned to Lorient's data under Le Bris so it may well answer a few of the questions I've been asking. Shows again the importance of considering data in conjunction with the eye test.
Yeah it's basically an 'introduction to Sunderland' article isn't it? Probably a bit early on to expect anything more in-depth I guess. If we maintain the results at the underlying numbers we've shown it'll be more worthy of a closer look, though.
Presumably teams will start looking to cut off those passing lanes from the centre backs to the full backs more. Will be good watching how we adapt, will obviously free up other lanes into the 6 or either of the more advanced midfielders or straight to the wingers - could force us to go a little more direct.
Are these the numbers you were referring to on your post on another thread (I think you were saying you couldn't find the post you were thinking of)?
Early signs from Isidor is that he's a footballer who likes to drop deep and get involved. Perhaps that could provide an alternative. Or maybe Mayenda can do it more as he comes up to speed. Not specifically - although they may have come from the same source. It was a table of where teams ranked against various metrics (xA f/a, xA, possession, field tilt etc) Think there were 8-10 metrics and each team was ranked 1-24. Was somewhat striking to note that our best stat was something like 3rd (from memory) and that we were midtable for a lot of others. Think it was Leeds, Boro and Coventry who had the broadest range of high rankings. Will be interesting to see if those rankings continue, if we or others regress to the mean or if it turns out that we are doing something a bit novel. Will see if I can dig it out. Can't for the life of me remember who posted it. Was on the Musk hellsite in case anyone else saw it.
From the basic Opta graph my simple conclusion is we are an outlier, that is our style of football is different from everyone else in the league at the moment. After that the devil is in the details.
Yeh thats what I was thinking - and that there must be some other graphic that shows why we're so good. The good thing is - that if Opta can't work out why Sunderland are good, it means other clubs data teams will struggle as well. Better to be an outlier that passes the eye test than predictable.
