Boards›[ARCHIVED] Fantasy LCS

Stats-base Projections - Week 6

Pseudo Q·7/4/2016, 1:44:53 AM·4 votes·1,109 views

[Edit] Updated to show how the results compared to the projections. [/Edit]

Week 5 was a good example of not relying on previous stats to try and predict the future. There were some major upsets, with teams like FNC failing to scrape even double digit kills over 2 games against the likes of GIA (who aren't a trash team?). Other teams are starting to clean up their play, giving away a lower number of deaths (where previously they were high) - IMT, UOL.

I've updated all of my posts from week 5 with a comparison of how the projections measured up to the actual results. It probably won't inspire a lot of confidence in being able to use my stats-based projections for future weeks (especially as I don't have as in depth stats as Riot), but week 5 certainly was an exceptionally odd week that few could have predicted the results for.

Week 6 Projections

Average Split Deaths/Kills = 11.53 http://i.imgur.com/GpgkMuO.jpg

** Week 6 Results**

http://i.imgur.com/EzhGHZQ.jpg

In an effort to help provide slightly more useful data to go by, I'm also going to add a couple of extra charts: Average deaths per week & Average kills per week. Rather than simply giving you a total average number of deaths and kills for the split so far, you may be able to identify which teams are consistent, which teams are altering their trends, and which teams are seemingly erratic (I would like to add in who each team faced each week, to add some context to their stats, but it's difficult squeezing it all into a single image).

Average Kills Per Week

http://i.imgur.com/B88q7jg.jpg

Average Deaths Per Week

http://i.imgur.com/hoRD7Wp.jpg

Like last time, I'm going to add a post for player position projections too (omitting the players that I don't have complete information on, because of substitutions, and would prefer not to fudge some numbers for the omitted players). But hopefully this week will be a bit more predictable... please!?

As ever, there is more to drafting than raw stats, but I will always appreciate upvotes for the effort.

8 Comments

Pseudo Q7/4/2016, 1:46:04 AM1 votes

[Edit] Updated to show how the results compared to the projections. [/Edit]

Week 6 Projections - Top

http://i.imgur.com/2vLARvP.jpg

Week 6 Results - Top

http://i.imgur.com/EszKSwi.jpg

Pseudo Q7/4/2016, 1:47:05 AM1 votes

[Edit] Updated to show how the results compared to the projections. [/Edit]

Week 6 Projections - Jung

http://i.imgur.com/FiEEQzF.jpg

Week 6 Results - Jung

http://i.imgur.com/BlpqAyg.jpg

Omitted

  • (FOX) kez
Pseudo Q7/4/2016, 1:47:49 AM1 votes

[Edit] Updated to show how the results compared to the projections. [/Edit]

Week 6 Projections - Mid

http://i.imgur.com/MFmuEmO.jpg

Week 6 Results - Mid

http://i.imgur.com/yKH5tL7.jpg

Pseudo Q7/4/2016, 1:48:44 AM1 votes

[Edit] Updated to show how the results compared to the projections. [/Edit]

Week 6 Projections - ADC

http://i.imgur.com/cLeKsE5.jpg

** Week 6 Results - ADC**

http://i.imgur.com/wtN9p1X.jpg

Pseudo Q7/4/2016, 1:49:28 AM1 votes

[Edit] Updated to show how the results compared to the projections. [/Edit]

Week 6 Projections - Supp

http://i.imgur.com/gtNIbND.jpg

Week 6 Results- Supp

http://i.imgur.com/T886OfZ.jpg

Omitted

  • (Cloud9) Bunny FuFuu
  • (Cloud9) Smoothie
Pseudo Q7/5/2016, 7:31:10 PM1 votes

I've just updated all of the above player projections to include almost all of the players that have ever been substituted. This factor was preventing me from being able to calculate a kill participation, and thus a projection of points for the week.

Now, I have extrapolated a kill participation via weeks where substituted players did play all games in a week, and ignored data (specifically for a kill participation calculation) from weeks where they were substituted part way through a week.

Players that have had their kill participation extrapolated have their name in red.

Oh, and I've also provided a Future value which adds the projections for all remaining weeks. Although someone may not be expected to get high points this week, it doesn't mean they aren't going to be a valuable investment for the future (provided you have the bench space for them). Conversely, some players may be disposable after this week.

Pseudo Q7/7/2016, 9:38:59 PM1 votes

I just done another update to the player position projections:

  • Added KDA.
  • Added historical data.
  • Refactored the formulas for substituted players (red names) so that I could provide historical data on them too, and discovered that I still had redundant formulas/statements from a previous version where I tried to fudge a projected value. So only those substituted players will have slightly different projections (and Future values) from before. mb! :)
Pseudo Q7/11/2016, 12:14:13 PM1 votes

Updated with comparison images for how the projections measured up against the actual results.

Week 7 projections will be coming soon.