Measuring Player Skill Follow Up Experiment

Do you mean combine each others scores or average 10 of my run and compare them to 10 of your run?

This sounds like fun. I generally autoplay most of the challenge event levels up to level 9. Rainbow team always; autoplay doesnā€™t care if a color hero is missing. Terrible scores bit I really just want to get them over with.

Might be fairer if everyone starts with the same team, but comparing the improvement delta over AI would be a reasonable alternative and much much easier to do.

Outside of Challenge Events, interested participants could also use a reasonably difficult map level.

This would make sense, and actually seems to be the best approach, if everyone had the exact same teams.

Edit: My understanding of your idea was that players rank themselves relative to each other on a standard level, using standard teams.

1 Like

My first data point of note

Iā€™ve completed 10 AI runs of 2-1-1-1

Having already learned from the initial results of you and @PeachyKeen I used this team

The AI completed 9 out of 10 runs successfully and the average score over those 9 was 44309.

(This included a dream final board where the bosses were annihilated in a cascade and I got a new high score!)

Iā€™ll complete my own ā€˜skilledā€™ runs tomorrow and see if I can beat you!

Continued

Iā€™ve now completed my ā€˜skillā€™ runs and have achieved an average of 51709.

So, firstly, thatā€™s higher than the AIā€¦phew

Dividing by my AI score, this gives me a skill index of 1.167 compared to @NPNKY score of 1.296.

So, the glorious nurse practitioner has me beat.

Can anyone claim the top spot?

2 Likes

This will be fun so letā€™s play, lol.

Firstly team used for mono in auto and manual mode was

Auto total scores where as follows. Won 8 out of 10 matches and no bonus rare heroes at all.

Enemies Defeated = 108,170
Time Bonus = 99,302
Match Bonus = 105,933
Health Bonus = 38,002
Total score = 351,407

In Manual mode total scores are as follows, won all 10 matches and received one Bonus hero/chest. No battle items of anykind used during any matches.

Enemies Defeated = 122,920
Time Bonus = 129,208
Match Bonus = 159,299
Health Bonus = 65,703
Total score = 472,130 with bonus score removed.

Will post the rainbow scores when completed.

Can we pick a less WE intensive stage? likeā€¦ 1-3? lol

OK this the 2.1.1.1 team used here

No healers or battle items used what so ever throughout this whole test and all battles where won in both manual and auto.

Manaul total scores where as follows. Won 10 out of 10 matches and no bonus rare heroes scores added here at all.

Enemies Defeated = 122,920
Time Bonus = 133,153
Match Bonus = 159,497
Health Bonus = 67,057
Total score = 482,627

In Auto mode total scores are as follows, won all 10 matches.

Enemies Defeated = 122, 920
Time Bonus = 126,010
Match Bonus = 139,804
Health Bonus = 50,265
Total score = 438,999

@NPNKY, Now that both of these are done maybe you can tell me where I sit in all of these stats compared to the others, lol, PLEASE.
Maybe @JonahTheBard can also calculate the score against the AI average as well.

It was fun.

I did almost lose 1 auto match Kvasir hanging in there with just 1 tile hit away from going but managed to get his mana up and win, just 98 H.B points on that game though compared to the average 5kā€™sā€¦lol.

All the rest of the matches without a healer managed quite well without any deaths at all. It certainly was an experience doing battles without using an battle items also, it changes the whole match.

I think the idea is to have the 3 bosses to make it somewhat challenging at the same time.

true, thereā€™s not much room to show off skill on 1-3ā€¦

but my WE x.x Iā€™m already running short, still not completely clear in epic top 500, and havenā€™t even made an attempt at legendary 1000 yet :frowning:

Morning @Ozy1

For your mixed team your average manual score, divided by your average AI score gives you a result of 1.099, which puts you in third, currently

For mono have you included the score of the runs that the AI lost? I canā€™t remember if @NPNKY included or excluded those

Youā€™re very welcome to have a go with a different level, @bobiscool, but you wonā€™t be in the running for Champion of the Universe.

I tried to pick a level that was as similar to a raid as possible, so I wanted the three bosses :see_no_evil::hear_no_evil::speak_no_evil:

1 Like

I just realized a way to ā€œgameā€ this ā€œgameā€;

set a team that is very likely to lose on auto-play; that way youā€™ll get 10 losses for auto for a very low average, and then you go in and beat the stage, voila, 3x or 4x score :smiley:

Oh no! :scream::scream::scream:

Yes all 10 matches (8 wins and 2 losses) are totalled into my results.

That sounds nice atm, lol. Which is better though, Ć  higher or lower score than this, I was never certain which way it goes.

Thanks heaps

@JonahTheBard, I am Readjusting these to only include the 8 won matches and the results are as follows.

Enemies Defeated = 98,336
Time Bonus = 99,302
Match Bonus = 88,735
Health Bonus = 38,002
Total score = 332,546

OK I have just completed another 2 auto mono battles making a total of 10 matches and the bottom results are a total score of 10 won matches of mono in AUTO mode.

Enemies Defeated = 122,920
Time Bonus = 125,584
Match Bonus = 115,113
Health Bonus = 53,342
Total score = 425,140

I hope this covers all possible results, :stuck_out_tongue_winking_eye: :rofl:

Cheers

1 Like

Just to clarify here as I may need to do a other 2 runs.

Is this test a total result of 10 win/lose runs or does it mean 10 Completed/won runs not counting losses in-between?

I included them all, win or lose

2 Likes

Okay, yā€™all, letā€™s refine things for next monthā€™s challenge event.

  1. Everyone uses the same team, with the following stipulations:
    ā€“all TC20 heroes
    ā€“no costumes
    ā€“no emblems
    ā€“all level 1 2* troops

We can work out the specific hero composition between now and then, but it needs to be the same for everyone, and it needs to be heroes everyone can access.

  1. We use the first triple boss stage on Rare, as before.

  2. 10 runs on autoplay to help establish an overall baseline. We take the average of everyoneā€™s autoplay runs together for the aggregate. Subtract any bonus points from chests, or remove and replay those runs.

  3. 10 runs played manually, with no item use. Screenshot your score screen from each run. Again, remove bonus points from chests or remove and replay those runs. Average your scores for those 10 runs.

Anyone who wants to participate is welcome.

@TheTimeHasCome, here is your engraved invitation

1 Like

Emblems are already allocated and I wouldnā€™t want to remove them just to put them back. They may also be needed for the raid Tornament which would be required with emblems.
Maybe alter that option please

If some of us have emblemed heroes and others do not, it invalidates the comparison. If we can come up with a consensus on a team, perhaps you could speed level a new copy of the heroes you need? 3* are cheap to level

So the next one is wonderland, I have at least one of every 3* with a few dupes, so letā€™s work on a team now then to make that happen, lol, Politely speaking of cause.

And yes I can speed feed and Max a 3* in a couple of days quite easily.

If you know what colours and bosses the stage that will be used will have it will help decide the heroes used.

1 Like