I'm so glad I found your channel. Very informative and helpful to me. Honestly, I wish I could find it earlier, because there are not many support only channels right here...
This video was extremely useful to me. I've played just over 6000 hours still stuck in Herald, have like 4 hero's at level 30 and keep blaming my team mates. This was eye opening, I'm certainly one who thought that grinding games was the only way. Thanks for the advice. Hopefully you do coaching!
I'm a huge fan of your videos, ZQuixotix. And this one is especially insightful. I find the point re. practicing outside of games (educational content, lobby mechanics, etc.) as key to improving quickly. Do you have any tips on how to split your time (games vs practice lobbies vs educational content & replays)? I also want to add an additional perspective when it comes to real-game practice - as long as you manage to maintain over 50% win rate, you should still play ranked instead of unranked. You are still gaining MMR. So as long as one is comfortable with a hero (and learning/practicing something else) and getting >50% win rate, I would recommend playing ranked. At the same time, leave the Chen practice for unranked. Thank you and keep up the good content! Btw, do you have any recommendations on a channel similar as yours, but for carries?
8:30 lmao A concept I think you forgot: if you're using the spreadsheet with large numbers, like 1000+ MMR, then in the later games, your win rate will naturally drop unless you keep improving, because you're relatively less better than your teammates and opponents. So the actual number of games is higher in practice. It's a bit discouraging once you do the math how long it takes to rank up. In my experience, playing pos 5, even if you're giga-smurfing, you can expect win rates of 60 (1k above) to 65% (2k above) at best. Going from ancient to immortal, then - even assuming you're already that good - will take literally months of grinding many games per day. That's a lot of toxicity to wade through...
Let W be total number of wins L total number of losses N total number of games -> N = W + L k the MMR gain from one win (+30 for solo) r winrate -> r = W/N = W/(W+L) and finally, G net MMR gain -> G = k*(W - L) G = k*(W - L) = k*(2W - W - L) = k*(2W - N) = k*N*(2W/N - 1) = k*N*(2r - 1) therefore *** N = G/[k*(2r - 1)] *** (the very formula he used in his spreadsheet) if we fix G at a desired value and plot N = N(r), it's a hyperbole! Unfortunately r can only vary between 0 (0% wr-> always lose) and 1 (100% wr -> always win). For any relevant MMR gain G, the plotted graph is still pretty vertical, meaning that: -no matter how good you are, it still takes time to gain a lot of MMR however -small improvements/losses in winrate will make a HUGE difference in saved/lost time Like all things mathematical, it's been under our noses THE WHOLE TIME but few people actually bother sorting it out on a sheet of paper (or of Google in this case). So thank you immensely for sharing it with us!
I'm issuing a copyright strike: ua-cam.com/video/IpEarhV4Ax8/v-deo.html Not really, but it makes me happy that there are more people interested in the actual facts and maths behind mmr. One interesting fact that always motivates me when grinding is that every win, all mmr you gain, is mmr that someone else loses. The average winrate is 50%, and around half of players have below 50% winrate. It's insane to think that if you have a positive ranked winrate, you actually have a winrate that's above half of ALL ranked players.
I love the tinkering with sheets and I've also made a "diary" of sorts where I can fetch my matches and check the MMR fluctuation. Upward trend is what keeps me going. Also the tips about playing ranked only when at "peak" performance is really good. Previously I've been playing it with friends with varying attitudes often leading to unnecessary losses.
4:50 This is dangerously close to gambler's fallacy--the coin has come up heads 4 times so I should expect it to come up tails the next 4 times in order for the observed probability to "even out." For it not to fall into gambler's fallacy, I would think you would have to demonstrate the person has gotten "lucky" and played themselves beyond their "true" skill level. I think that begs a lot of questions about what it means to be lucky and what true skill is, etc., etc. I loved that you pointed out that pulling a win out can save you hundreds of hours to your MMR goals (potentially).
"More to life than MMR, more to life than DOTA"
Questionable.
I'm so glad I found your channel. Very informative and helpful to me. Honestly, I wish I could find it earlier, because there are not many support only channels right here...
Better late than never!
I went on an 18 game win streak - never felt more anxiety about queue for dota. hate grinding mmr.
This video was extremely useful to me. I've played just over 6000 hours still stuck in Herald, have like 4 hero's at level 30 and keep blaming my team mates. This was eye opening, I'm certainly one who thought that grinding games was the only way. Thanks for the advice. Hopefully you do coaching!
I can't wait to apply this to my Wisp game.
Thanks for sharing the sheets!
Dont mind me, just your daily algorithm- enchancing chicken burrito 🌯
The best kind of chicken burrito
Great guide as always Z, best Dota content out there right now!
No way you actually play(ed) OSRS, what a legend!
Keep up the good content !
Really helpful videos 👊🏾
Great video as always ❤❤
I'm a huge fan of your videos, ZQuixotix. And this one is especially insightful.
I find the point re. practicing outside of games (educational content, lobby mechanics, etc.) as key to improving quickly.
Do you have any tips on how to split your time (games vs practice lobbies vs educational content & replays)?
I also want to add an additional perspective when it comes to real-game practice - as long as you manage to maintain over 50% win rate, you should still play ranked instead of unranked. You are still gaining MMR. So as long as one is comfortable with a hero (and learning/practicing something else) and getting >50% win rate, I would recommend playing ranked. At the same time, leave the Chen practice for unranked.
Thank you and keep up the good content!
Btw, do you have any recommendations on a channel similar as yours, but for carries?
Inspirational content!! Self-promotion skills 💯
Very useful when you wanna get gud
8:30 lmao
A concept I think you forgot: if you're using the spreadsheet with large numbers, like 1000+ MMR, then in the later games, your win rate will naturally drop unless you keep improving, because you're relatively less better than your teammates and opponents. So the actual number of games is higher in practice.
It's a bit discouraging once you do the math how long it takes to rank up. In my experience, playing pos 5, even if you're giga-smurfing, you can expect win rates of 60 (1k above) to 65% (2k above) at best. Going from ancient to immortal, then - even assuming you're already that good - will take literally months of grinding many games per day. That's a lot of toxicity to wade through...
Not really relevant as your not likely to watch these videos if you are 1000k MMR player.
Thank you
Happy new year to you and everyone here! o/
Happy New Year!
Let
W be total number of wins
L total number of losses
N total number of games -> N = W + L
k the MMR gain from one win (+30 for solo)
r winrate -> r = W/N = W/(W+L)
and finally, G net MMR gain -> G = k*(W - L)
G = k*(W - L) = k*(2W - W - L) = k*(2W - N) = k*N*(2W/N - 1) = k*N*(2r - 1)
therefore
*** N = G/[k*(2r - 1)] *** (the very formula he used in his spreadsheet)
if we fix G at a desired value and plot N = N(r), it's a hyperbole!
Unfortunately r can only vary between 0 (0% wr-> always lose) and 1 (100% wr -> always win). For any relevant MMR gain G, the plotted graph is still pretty vertical, meaning that:
-no matter how good you are, it still takes time to gain a lot of MMR
however
-small improvements/losses in winrate will make a HUGE difference in saved/lost time
Like all things mathematical, it's been under our noses THE WHOLE TIME but few people actually bother sorting it out on a sheet of paper (or of Google in this case). So thank you immensely for sharing it with us!
This guy maths!
as someone who reached the bottom at one point.. you can't actually reach 0 mmr. it just stop dropping at 1mmr.
OSRS shout out yoooooooo.
Great videos man! PMA
More videos like this please bruh.
I'm issuing a copyright strike: ua-cam.com/video/IpEarhV4Ax8/v-deo.html
Not really, but it makes me happy that there are more people interested in the actual facts and maths behind mmr. One interesting fact that always motivates me when grinding is that every win, all mmr you gain, is mmr that someone else loses. The average winrate is 50%, and around half of players have below 50% winrate. It's insane to think that if you have a positive ranked winrate, you actually have a winrate that's above half of ALL ranked players.
Is the winrate only in ranked?
Yeah, only your ranked win rate will matter for gaining MMR
I love the tinkering with sheets and I've also made a "diary" of sorts where I can fetch my matches and check the MMR fluctuation. Upward trend is what keeps me going.
Also the tips about playing ranked only when at "peak" performance is really good. Previously I've been playing it with friends with varying attitudes often leading to unnecessary losses.
Basically theory told I need to win. Your theory is good. But it is planning.
People like me playing 0-2 ranked games a day, and keeping 70-80% winrate, feels pissed off after every loss.
4:50 This is dangerously close to gambler's fallacy--the coin has come up heads 4 times so I should expect it to come up tails the next 4 times in order for the observed probability to "even out." For it not to fall into gambler's fallacy, I would think you would have to demonstrate the person has gotten "lucky" and played themselves beyond their "true" skill level. I think that begs a lot of questions about what it means to be lucky and what true skill is, etc., etc.
I loved that you pointed out that pulling a win out can save you hundreds of hours to your MMR goals (potentially).
50% of players will under 50% win rate though
Hahaha i also played tons of osrs
I don't have enough time for both games rip