Random To Do List (RTDL) Suggestions Thread


#41

I don’t think I’d want more the 25, personally. Except maybe as a one time blowout as T suggested. But with 25 I feel there is always an outside chance I can get them all without having to totally neglect other gaming. Also going over 25 would give even more of an advantage to folks with more free time (for those who are trying to compete for the top spots).

Giving 3 or 4 months a go sounds good to me.


#42

What about something like this?

Assume your maximum total points are 1000 and your achievement ratios are (1.08, 1.42, 2.26, 2.54, 3.36). I only chose 5 for simplicity.

First, find out what % of each achievement is of the total ratio. Then use that % to figure out it’s points. So, here’s the point values for those ratios:
Total ratio = 10.66

  • 1.08 is 10.13133208255159% of the total so it is worth 101
  • 1.42 is 13.32082551594747% = 133
  • 2.26 is 21.20075046904315% = 212
  • 2.54 is 23.82739212007505% = 238
  • 3.36 is 31.51969981238274% = 315

That totals to 999 because of rounding. To fix that, just tweak the points for one of the achievements… so like make the first achievement worth 102 points.


#43

Sure, I think this is essentially the same as my suggestion of normalizing to the same maximum point value. I suggest finding 1000/1066 = .9380863
Then multiply all your point values by .9380863.
Thus 108 * .9380863 = 101.3133204.
etc.
Same values as you got.


#44

Yes, it’s two different ways to calculate the same thing. I’ll point out that anything that requires manual tweaking will not be considered, this needs to be automated and automatically calculated as much as possible.

While I do like the concept of weighted scores in the interests of evening up the field, I do have concerns:

  1. It removes an element of the randomness, namely the value of your list. Part of the random nature of the contest was that the total value of your pool could come out to a range of values - normalizing it in this way completely removes that element.
  2. It could be abused. If someone floods their entire pool with very low ratio achievements they are not penalized in any way - their pool will have the same value as another with many high ratio/rare/difficult achievements.
  3. It’s complex. While simplicity is not absolutely necessary, I feel that a competition that has calculations that are easy to understand at a glance makes it more approachable.

#45

The harder/longer achievements are usually the 1.3 finish the 20 hour campaign ones! Haha.


#46

Having access to our old lists would be welcomed. I can’t remember why I wasn’t into this contest the first 2 months.


#47

I’ll be making this happen during August, once I’ve had time to test out a bunch of other improvements and ideas I have brewing.


#48

Ok, so I’m pretty sure the next iteration of this competition will be as follows:

  • A tri-annual seasonal format, with each season being 4 months. The seasons would be January to April, May to August, September to December.
  • The number of achievements will remain at 25. This seems like a reasonably challenging amount for most people - it’s possible, but not generally easy. Going higher than this also raises further questions about the size of the draw pool and stacking lists.
  • Minimum available pool size will be set to 1,000. This value is a balancing act, as going much higher than this will actively penalise people with high completion percentages - and those aren’t the people we’re really targeting with the minimum pool value. 1500 seems to be a rough (eyeball only) average for people with mostly unadjusted pools, so 1,000 gives a bit of wiggle room.
  • The ranges of the draws to be adjusted as follows, where the pool is sorted by ratio descending (that’s just the way Rich has set it in the API):
    • 14 achievements from the 50-100% range (the lowest ratios)
    • 7 achievements from the 25-50% range
    • 3 achievements from the 10-25% range
    • 1 achievement from the 0-10% range (the highest ratios)
  • A maximum draw value to be enforced. The value of this is still to be determined as I need to run a pile of test draws with the new ranges to see what ends up being the average.
  • If your draw is below a certain value (to be determined) you can ask for a complete re-roll of your list. This is to avoid falling behind too far because of a bad draw.
  • IF I can get some time with Rich I plan to add the following per-person options:
    • DLC (but NOT Title Updates)
    • Select from started games only

I have had another idea that I’d like to throw out there: Bonuses. If you qualify for a particular bonus, you would get bonus points (value to be determined).

Examples of bonuses might be:

  • Unlock 10 achievements in ratio order, ascending.
  • Unlock the top 5 highest ratio before any other in your list.
  • Unlock 10 achievements in alphabetical order (either direction).
  • … more?

Let me know what you think of the bonuses idea and if you have any good ideas for recurring bonuses.


#49

I like the bonuses idea


#50

I like the bonuses idea too - I’m having a blast figuring out how to schedule my list to fit in with Leap Frog, and that would be similar. I’m assuming they would all have to relate to game or achievement name, TA value, TA ratio, or GS, since everything else will vary by person - would it be one bonus a month, or would they all be active at once?


#51

I’m thinking they would all be active at once, although we could also throw in random bonuses that are harder each month.

I’m also thinking that some of the bonuses should be exclusive - i.e. if you go for one bonus, it will lock you out of getting another just by definition. It shouldn’t be possible to get every bonus in a month. For example, having a bonus to “unlock your 10 lowest ratio achievements in order from lowest to highest first”, and another bonus for “unlock your 5 highest ratio achievements first” would be exclusionary as you can only do one of them “first”.


#52

Looks good to me!


#53


#54

Looks like fun, but a lot of work. Thanks, dude!
The bonus idea sounds great. And you could probably add them as you go (month to month).


#55

I like the idea of bonuses too. Kinda like GTASC where you jot down a few and pick one at random for the month?


#56

The idea was modelled off GTASC, so randomly selected bonuses was also on the table if we could think of enough.

Given my home situation and how close we are getting to the new contest start I’m going to have to put the bonuses idea on hold, at least for the first month. I’ll be struggling to get everything all working in time as it is.

It’s going to be awesome though - I’m adding things like tracking the daily leaderboard movements and timestamping when achievements are found, which can make some interesting data comparison or graphs later. I’m also adding more robustness to the notifications (the process that posts the unlocks in the thread here) so that it will retry ones it misses.

I’ve basically been rewriting the whole system from scratch; the original was pretty cobbled together, and I’m doing it properly this time with a common library and tests. I have the achievement selection and achievement checking parts working now, and just need to write code to handle the ranking (going to do most of it in code rather than use the spreadsheet now), updating ratios, and notifications. I then need to put some thought into how the historical information is going to work.


#57


#59

Curious if possible - when checking monthly leaderboard, clicking on someone’s gamertag taking me directly to their list without having to scroll through all the tabs on the bottom.


#60

I’d looked into that in the past - unfortunately it’s not simple, but I have more control over what is put on the leaderboard page now (nickname support) so I might be able to do something there. I’ll put it on the list, thanks for the reminder.


#61

This may be obvious, but I didn’t realize it at first. Use the double hamburger on the bottom left. It is easier than scrolling through the tabs.