NA Statistical predictions with Spellsy's Stats Supershow!
Predictions for NA Superweek using a statistical model based on this split's records and recent matches.
This article was originally published on GameSpot's sister site onGamers.com, which was dedicated to esports coverage.
While the NA standings are not as close as the EU standings, there should still be plenty of drama going into superweek. I built a model to create predictions on this weekend’s outcome, and the results are as follows:
![No Caption Provided](https://www.gamespot.com/a/uploads/original/1546/15461225/3036743-ogImport-7999-na%20standings%20list.png)
The model uses a team’s win rate on blue or red side, the matchup’s head to head record (ex: since C9 has a 2-1 record against TSM, they have a “66.6% win rate” against TSM, while TSM has “33.3% win rate”), and I looked at the win rate of their last 5 games (ex: TSM has gone 4-1 in last 5 games, giving them an “80% win rate”). I then took these three factors and added them together, giving double weight to the blue/red side statistic, as that is a very important stat. After calculating each teams likelyhood to win, it is then compared to each other, and the team with the higher value is predicted to win. I then used this difference to create a “strength of prediction” seen below. An example here can be TSM vs C9. TSM has a .916 blue win rate, .8 win rate in last 5 games, and .333 h2h win rate, C9 has respectively .6666 red win rate, 1.0 win rate in last 5 games, and .6666 h2h win rate. Putting this together (given that blue/red win rate has double weight), TSM has a prediction value of 2.97, while C9 has a prediction value of 3.00, giving a 50.4% likelyhood of C9 winning.
Here is the stats breakdown of each match of the superweek:
![No Caption Provided](https://www.gamespot.com/a/uploads/original/1546/15461225/3036744-ogImport-8002-na%20predictions.png)
As you can see, there are some very close matches such as CLG vs C9, XDG vs DIG, TSM vs C9, EG vs CRS, and XDG vs EG. I wanted to break down in more detail a few of these matchups
The race for first place
There are two big battles here for key spots, one is the battle for first place betwen TSM and C9. They will most likely be decided by two key matches, TSM vs C9 and CLG vs C9. The important factor here is that C9 will be on red side in both of these key matches. C9 has a 100% win rate on blue, with only a 66.6% win rate on red - this is a huge difference, and while C9 is on a huge win streak (8 in a row) it is still a nervous week as they play both of the other top tier teams with a significant disadvantage.
The race for last place
The second big battle is the race for last place, which is very important due to the fact that LMQ is lurking in the challenger scene, and you don’t want to have to face them. Currently XDG and EG are tied, and they play each other for the last game of the season. This could be a really huge game seeing as the predicition model doesn’t have either of them winning a match up to that point. Again though, XDG is at a big advantage because of their blue side, they have won 5 games on blue while only 1 on red - it is clearly their best side and they almost have a winning record on blue (5-6). Overall, blue vs red will be a huge factor this weekend as we see the closing of the split.
Data collected with help by the onGamers Stats team: Derek 'Kathix' Adams, Steven 'whedgehead' Falgout, Kent 'Traepoint' Frasure, Jake Morales, and James 'PelkaSupaFresh' Pelkey. Design by Ben 'Sarcasmappreciated' Li.
Got a news tip or want to contact us directly? Email news@gamespot.com
Join the conversation