1. It makes it easier to SEE YOUR IMPROVEMENT TRAJECTORY (how fast are you improving), based on the month's ladder MMR data, with some DEGREE OF ACCURACY and PRECISION, INSTEAD of A VAGUE ABSTRACTION,
2. It makes it more FUN and adds a FUN CHALLENGE, in that it's VISIBLE WHO IMPROVED their ladder MMR * difficulty weight THE MOST DURING THE MONTH (essentially, who was improving most effectively in the past month or, in other words, has the best improvement trajectory from the past month's data)
Important note - slight adjustment of the MMR difficulty multiplier:
WHEN:
- applied for JULY 01 - 31 and, save for specific value, probably here to stay
WHAT:
- in the ladder (month's end MMR - month's start MMR difference) * difficulty weight, I will heighten difficulty weight (multiplier) for above 2k ladder MMR's
- currently, for month's start 1850 MMR it's difference * 1,850, for 2224 MMR it's difference * 2,224, and so on
HOW:
-by how much - unknown but it will be kept as simple as possible
FOR WHAT:
- it's first about symbolically expressing that it's quite substantially harder to improve above 2k than below 2k, even if it's 1800, 1900 MMR - it becomes a much, much steeper territory to climb
- and then, second, it's about adjusting the specific value of the MMR difficulty multiplier to be as accurate as it can be
- it's imperfect and will stay imperfect
- the point isn't to make a perfectly accurate representation, in the same way that the point of wearing a watch isn't to have an atomic watch that would enable being perfectly on time for a meeting in terms of nanoseconds
- the point of this adjustment- and slight adjustments that will likely follow in the next months - is to reach a satisfactory and practical threshold in terms of accurate representation that matches the realities of laddering and MMR reasonably well
MONTHLY LADDER IMPROVEMENT CHAMPIONS - HALL OF FAME:
MAY WINNER - Red_LML.
- IMPROVEMENT SCORE: 360.
- IMPROVEMENT DATA: 1800 MMR to 2000 MMR.
JUNE WINNER - GruGloG[ReV]
- IMPROVEMENT SCORE: 11.
- IMPROVEMENT DATA: 1514 MMR to 1525 MMR.
JULY WINNER - ...
AUGUST WINNER - ...
SEPTEMBER WINNER - ...
OCTOBER WINNER - ...
NOVEMBER WINNER - ...
DECEMBER WINNER - ...
HOW DO I PARTICIPATE?
Just open google spreadsheet and provide:
1) username,
2) ladder score from the current month's start,
3) ladder score from the current month's end,
Only one note, and you don't even need to read that - unless you want to add anything other than just MMR in the MMR cell - but if you do, please add it in the cell's comment (such as something about ladder reset etc.). As a side note, for now, issues such as ladder resets, are not considered because the base idea has to be proven solid, first).
That's all.
Monthly ladder improvement - google sheet:
docs.google.com
WHAT IS IT?
So I had this idea to make laddering more interesting. It's very simple.
Monthly winner is the one with the highest ladder improvement score.
The ladder improvement score is: (next month's ladder mmr - current month's ladder mmr) * MMR difficulty weight.
Calculation formula examples & details are provided in the google sheet column header comment, the first of ones titled IMPROVEMENT. It's actually really straightforward. No crazy math in there.
Participation criteria - just fill your data in the linked google sheet. It's set so anyone can edit. Nothing more is required. I want to keep it as simple as possible.
First day of each month, the person with the highest improvement score on the google sheet is verified on battle.net and gloriously declared the monthly ladder improvement race winner.
Ladder improvement score = next month mmr - current month mmr * weight. Weight is just decimal from current month ladder mmr.
The difficulty level weight is there because it reflects the nature of improving. Challenge & difficulty of climbing Mount Everest's 1st kilometer vs climbing 5th kilometer is naturally very different, and climbing 5th kilometer vs 8th kilometer is pretty incomparable.
Details:
+ Show Spoiler +
IMPROVEMENT = LADDER MMR month(n+1) - LADDER MMR month(n) * WEIGHT
WEIGHT = LADDER MMR month(n) TO DECIMAL POINT, E.G:
2500 = 2.5
2000 = 2.0
1800 = 1.8
1400 = 1.4 etc.
EXAMPLE:
MMR 1000 to 1500 = (1500 - 1000) * 1
IMPROVEMENT SCORE = 500
MMR 1500 to 1725 = (1725 - 1500) * 1.5
IMPROVEMENT SCORE = 337,5
MMR 2000 to 2225 = (2225 - 2000) * 2.0
IMPROVEMENT SCORE = 450
Over time, I'll add analytics relating to what's really interesting - improvement rate. That is comparing one month's improvement vs next month's improvement trends. This can get really interesting. Think graphs representing improvement per username, increase of the improvement rate per username.
I've also had the following ideas:
- adding best guide of the month backed by results competition - with the focus on how to accelerate one's improvement rate
- adding best guide about accelerating one's improvement of the month voted by the community
- monthly ladder improvement race winner playing offrace only
The basic idea of this is similar to speedrun's, except instead of finishing the game in the least amount of time possible, the goal is to improve one's ladder MMR by as much as possible.
That's just the basic so there's something concrete to work around.
However, it's completely about the improvement rate. If in month 5 to 6, you figured out how to improve by 1.5x of what you improved in month 4 to 5, and you repeated similar trend in the next period, now that's the real win. That's an incredibly difficult and amazing accomplishment.
For now it's just a basic test but over time, if there'd be interest, I think it's very interesting to work with the ideas above and see how people perform.
For participants, about score calculation:
I want to remind that this is an initial test. I will try to find a way to fine tune the method to calculate improvement score with enough accuracy. So as to account for the realities of how the better you get, the harder it is to get even better. Right now, around 200 MMR improvement below 2k can amount to > 300 improvement score. Around 100 MMR improvement above 2.1k MMR, can amount to < 300 improvement score. The MMR weight's in need of some data, testing & fine tuning to get it fairly representative. (Note date - beginning of June)