Behind the Numbers: Passing and Winning

I have a love for numbers and statistics. It is probably for this reason that I chose to have it as a major part of my career. Because of my interest in learning more and using statistics to find supporting evidence, I have enjoyed incorporating them into articles I write here.

A conversation last week led to the decision to finally begin putting together a regular series on different statistical aspects of the Vancouver Whitecaps. An attempt to see whether what we see (or think) is supported by the numbers. The first step in this process was examining the long-discussed question around whether possession was important to success (measured by points). I won’t spoil the conclusion for you, but will provide you with a link to the article.

While the first measure in determining what statistics are useful for predicting success was possession, the next is passing accuracy. Like with the possession analysis, I will use data found on the MLSSoccer.com website and WhoScored.com. In my article on possession, (former 86Forever editor) Rob R. Scott noted “Russell Teibert can pass all day long, and complete 93% of his attempts. Unfortunately, the vast majority of his passes are simple pass backs…”. As a result, I am going to take his advice to try and break down passing a bit beyond accuracy. However, it will be my starting point. Like with the previous article, I will measure success by points. Feel free to suggest a different measure of success.

Does Increased Passing Accuracy Lead to More Success in MLS?

Unlike the previous article, I will not go into too much detail about the general statistics here. However, I will note the correlations between passing accuracy and success over the past three seasons. In 2015, passing accuracy was slightly correlated with points achieved at home (0.08) or away (0.13). In 2016, the numbers were similar (0.09 at home versus 0.17 away). In the social science world, those numbers are quite low. This season the story has been different. Overall, passing accuracy was correlated with more points per match at 0.36. This was entirely because of home matches, with the correlation there being 0.27, compared to away where it was -0.01. Will this pattern hold? Previous season numbers would suggest no. Nevermind no correlation, but in 2015 and 2016, passing accuracy was more important (though still not very) on the road than at home.

What conclusions do you draw from this?

Where do the Whitecaps Rank in Types of Passes (2017)?

The next step was to examine different types of passes. I will focus on 2017 and then we can look at any interesting numbers that arise, in previous seasons. No surprise, but MLSSoccer.com was very poor in providing statistics. So, this analysis will rely on numbers obtained from WhoScored.com. To have a look at the numbers yourself, go here. However, I will provide some of the main numbers I used.

First, I will note that the total number of “Long Balls per Game’ and ‘Short Passes per Game’ do not add up to ‘Total Passes’. Why? Not sure. But, those are the numbers that were provided, so those are what I will go with. Second, these numbers are for overall. I will provide a summary of Home and Away, but won’t include the tables.

General findings first. I think it comes to no one’s surprise that the Whitecaps are last in total passes per game (339.3). This is more than 20 less than the next closest, and 2/3 the amount of leading Sporting Kansas City. Second, breaking the Vancouver Whitecaps down further, they lead the league in Crosses PG (20). Combined with their fewest passes per game, that means the Whitecaps are the most reliant on crosses, accounting for 6%. The next closest is DC United at 5%, Real Salt Lake at 4.7%, and Houston Dynamo at 4.4%. Noticing a pattern with those team names?

Third, the Caps are middle of the pack in number of long balls, but near the bottom in accuracy (43.6%). Who is below them? Houston, Salt Lake, DC United, and, surprisingly maybe, New York Red Bulls. Fourth, while completing the fewest short passes per game, the Whitecaps are only ahead of DC United and NYRB in accuracy of short passes.

To summarize, in 2017 the Whitecaps average fewer passes than anyone else, are one of the worst in completing passes long or short, and rely on the cross more than any other team, despite their main striker (Fredy Montero) being 5’9”, and their wingers being 5’9” (Christian Bolanos) and 5’2” (Cristian Techera). For what it is worth, Brek Shea, who has played striker a bit is 6’3”.

Do Certain Types of Passes Lead to More Success in MLS (2017)?

Next, we will correlate some passing statistics with the number of points per game in 2017. I wasn’t able to get information on types of passes without going to each individual match and I wasn’t willing to put that type of effort in for this article. However, I did look at Total, Long, Short, Proportion Cross, % Accurate of Long and Short, Total Key Passes, and Long/Short Key Passes. I broke it down by Overall, Home, and Away.

Some general findings. Relying on crosses at home was VERY bad, but inconsequential on the road. In fact, it was probably the best measure and if I were to go into some regression analysis, I would probably use that as my first predictor. Being accurate on long balls on the road and short balls at home were correlated with success. On the road, relying on long key passes was bad while relying on short key passes was good. This is kind of surprising, given that (accurate) short balls at home were good while (accurate) long balls away were good. There are plenty of more interesting findings in this data, when you start mixing and matching the combinations (as I just did), but I will leave you to pick out which of those you found most interesting. Let us know in the comments.

Where do the 2017 Whitecaps Rank?

You may be curious how the Whitecaps breakdown on the various passing statistics. I covered some above, but I will provide a bit more detail here. The Whitecaps rely on the cross more away (6% vs 5%), however, they rely on crosses more than any other team at home (remember above I said doing that was REALLY bad).

Their long ball passing and accuracy is the same home or away, They are fourth worst away from home in long ball accuracy. Combined with averaging the fewest short key passes, road success does not seem likely going forward (remember, long ball accuracy and short key passes was correlated with road points).

While the Whitecaps average slightly more short passes per game at home than away (291 vs 260), they complete approximately the same percentage (80.3 vs 78.8) regardless. Again, recalling that short ball accuracy is correlated with success at home, the Whitecaps are, again, fourth to last.

Finally, when the Whitecaps win this season, they average fewer total passes (319) than when they draw (329) or lose (348). However, their passing success is the same (73% in wins, 73% in losses, and 69% in draws).

OVERALL CONCLUSIONS

From all of this data, what can we conclude? Well, in general, there are some passing stats that are important for success, but the location of the match (home or away) may dictate which are important. As for the Whitecaps. They suck at passing and the types of passes they do the most (crosses) is the least likely to lead to success in MLS. This is amplified by the short team the Whitecaps have, going counter to their skillset. I could go on, but I feel I would be too negative. Instead, I will leave that to you. What conclusions do you draw from all of this data about the Whitecaps and about MLS?

P.S. If there are certain numbers you are interested in looking at more closely, let me know in the comments and I will add a table with the data.