Training Camp lvl 20 - Results

That is certainly true, though it doesn’t mean that the results are not random.

We would need large samples from at least two camps to test whether the underlying chance of each producing a 5* hero was different. How large? I haven’t worked it out and I don’t plan to, but my instinct says that 1000 trainings from each camp would probably not be enough. Probably not nearly enough.

Haven’t paid attention to this in a while - see the counting has stopped (and fair enough).

In that case I will just complain that I am now 3 out of 123 and it’s been 48 pulls since my last 5*. This total puts me on less than half of the average I guess :frowning:

Ah well, can only mean my luck will eventually change for the better :slight_smile:

1 Like

youre doing better than me. im 2 of 102 and it’s been 70 pulls since my last 5*

3* - 2 (Greymane today)
4* - 1
5* - 0

Stop the counter bro, it carries you bad luck, I know from my experience, and good luck!

I’ll edit this post periodically.

5* - 0
4* - Scarlett
3* - Friar Tuck

My second 20- now running two at 20 - was spot on the average
1- 5* Elkanen
4-4*
15-3*
Keeping my fingers crossed for the elusive 5*yellow

Fifteen TC20 attempts so far, and no yellows of any level at all for me. Good luck.

Odds of that are about 1:3. So not surprised.

More like 1 chance in 30 if I recall my maths correctly. At this number of results, just a statistic curiosity though. If it holds up for a couple more weeks then I might start wondering if there’s something odd happening.

1 Like

Interesting. I guess we’ve never considered whether the odds of drawing different colours might vary. We could collect some data. Does anyone care?

I’ve just kept the color figures out of idle curiosity. It wasn’t any extra trouble with a spreadsheet. I figure they’ll even up as I get more results over the longer haul.

Edit: Got Chao today, so the no-yellow run is over after 16 total of other colors. Hopefully, the non-5* run will end soon too.

i’ve posted the odds of color a few times in this thread, though i didn’t explicitly call that out. we’re tracking that on the spreadsheet. here’s the latest numbers. while it’s not broken out by 3,4,5* here, the raw data exists for that if it’s ever of interest.

2018-01-21_2315

1 Like

Since my last post

4x 3*

Personal Total

39x 3*
11x 4*
0x 5*

*sigh hopefully the 3rd tc20 I’m working on is more lucky…

Especially useful as defense in the alliance wars.

3* - 3 (Valen today)
4* - 1
5* - 0

Struggling to keep this on topic, so I’ll start with this: a math geek I tend to think in terms of ratios not percentages, so the statement earlier in the thread that people would tend to go with round numbers over something like 19%, 6%, 75% doesn’t really ring true for me. I personally would start out thinking to myself what ratio of heroes ought to be 5 star vs 4 star vs 3 star and in this case 1 out of 16 would be 5 star, 3 out of 16 would be 3 star and the remainder would be 3 star (this is an example of what I would do as a programmer folks). That would mean the percentages would work out to be 6.25%, 18.75% and 75%.

As far as Leonidas goes, (Ha! Kept it on topic) He’s worth leveling up as he is good in different situations than Jackal and much sturdier fully trained. He’s stronger than Chao, so feel free to spend the items on him IMO.

1 Like

I mean, I may level him so that I can you use him on titans and yell ‘THIS IS SPARTA’ when the battle starts

:slight_smile:

I have some other folks to level first to get ready for events and what not, but I’m going to consider it for sure if I don’t get something better.

This logic is compelling. Put differently:

  • 1 in 4 draws will be epic or legendary
  • of those, 1 in 4 draws will be legendary

The values (1/16, 3/16, 12/16) are comfortably inside the confidence intervals. I think we have a winner.

1 Like

Never thought of it that way: it’s an interesting way of looking at it. I don’t know whether you’re right, but you’ve definitely convinced me that you might be.