Interesting that you'd point at the missed tackles at Attwood. There was an article by Wayne Smith a couple of years which is pretty much engraved on my mind where he talks about the fallibility of stats and his prime example is indeed the missed tackle. We look at the stats, see a missed tackle and assume a mistake was made; yet he said that a missed tackle that makes it easier for a team mate to complete the tackle before the gain line is a more positive intervention than a successful tackle that is finished behind the gain line. Over time I've come to agree with him and all the other proponents of it and wouldn't be weighing missed tackles that heavily.
True, stats are fallible, and not all missed tackles are created equal. Presumably though some missed tackles are not all beyond-the-gain-line-slow-downs and are just good old misses. You've always got to watch the game. What we've found is stats can throw up some anomalies if you just look at a single game, but over time they are a reliable gauge of talent. If someone consistently has a number of missed tackles, even Wayne Smith might have something to say about it.
So, the team above is just one of several plausible lineups that you could put forward based on just the first round of games, since the data sample is so small. As the weeks progress we think we'll be able to zero in on the real standouts.
Also have to say, I think it rather unfair that you'd weight towards the winning team; generally, the guys on the dominant team get more chance to shine anyway, which already weights things towards the winning team. To weight it further is unnecessary and unfair on players whose teams simply aren't going to win that game. Do you weight based on form/rankings as well?
Yes, we do adjust ratings based on the ranking of the protagonists. That's usually league standing, or in test rugby we use the World Rugby rankings. So, we rank a performance in a game of 1 vs 2 higher than an identical performance in a game of 11 vs 12. It's probably one of the most hotly debated aspects of our ratings, but if you consider the alternative, then two tries for Fiji against Japan would be ranked the same as two tries for South Africa against the All Blacks.
So in the case of this round of the 6 Nations, players in the Wales/England got the biggest lift, although as you can see from our list, it's not a crazy bump, and players from other matches can make this list. The weight towards the victors is also something that can set people off. If it helps, margin of victory is a factor, so if your team goes down 14-15, then any bump to the winners in nominal.
Anyway, we do agree that stats will sometimes need to be taken with a pinch of salt, but if you want to win an argument down the pub, it's always good to back it up with some objective metrics. That is, so long as they support your viewpoint
Thanks for taking a look, and thanks for your thoughts.