Good video the only thing I have to add is that the winrate of items are not exactly that easy to compare when there is a large pick rate difference. An example that phreak had was that stating boots no pots on hecarim jungle one season had like a 5% higher winrate because all the players that stated with no pots were hecarim mains and won more games regardless of the starting item, according to phreak he estimated that you should assume the winrate of an off meta build will decline by about 2% if it became the meta build. So you still have to do your own testing and conclusion of the items
Thank you for your feedback, i appreciate it! Yeah, i agree, stats can be hard to read sometimes, i'll do a video about it someday. There are 2 important things about items winrates that you need to keep in mind. 1. is amount of games that was played with certain item. 1-2k games is the minimum to judge an item, but it can fluctuate a lot, so it's better to use larger sample size(30 days)/track an item for many patches/use gold+ stats. 2. gold cost. Low cost items should have lower winrate than higher cost items. However, when low cost item (for example statikk in Jhin's case) has more winrate than high cost item (collector), it's pretty clear which item is better. Not sure about Phreak statement tho. I think he accounted for the margin of error in his conclusions, but the margin of error depends on the number of games played. For example, i won't even consider to buy an item with 1-2k games if it has +-the same winrate as the most popular item, because winrate in 1-2k games can fluctuate a lot. I recently found a cool example of how pickrate doesn't mean that much so let me share it with you. ADC Brand got some popularity not so long time ago, so he's the perfect champion to check how pickrate affected his winrate. lolalytics.com/lol/brand/build/?lane=bottom&patch=14.8 Brand adc stats for 14.8 patch: 55.76% winrate, 9k ....a couple of patches after.... lolalytics.com/lol/brand/build/?lane=bottom&patch=30 30 days stats, 80k games, 54.92% winrate. Worth to mention that he also got nerf between 14.8 and last 30 days, so overall, his winrate was almost +-the same, if we take into account 1 nerf.
@@armasot nah its not "margin of error" but the general pattern that players that alter their build will on average be also just better on the champ (since you generally only deviate from the common path, if you have an idea _why_ you'd deviate); this was shown with Items (Trinity vs Divine Sunderer) in S12 on a champ (which I sadly forgot); where swapping which is recommended raised the winrate of the "standard" item and lowered that of the "sleeper OP" item. They replicated a similar effect lately with runes, where most of the time the new recommended rune lost about 2% winrate and the no longer recommended one gained about 2%.
Well, i think it depends on champion's skill expression. Champions with high skill expression will lose winrate on any recommended items and runes for about 2%, because lower elo players won't be able to play this champion no matter what (checked it with Azir change, coup de grace to cut down). For champions, who is just average in any elo this % will be much lower or won't exist at all, like with Trundle top (pta was changed to grasp as recommended, didn't lose any winrate or Brand adc, who got more popular recently, also didn't lose any winrate up until now with nerfs). Also, with Trinity vs Divine Sunderer, i think it was Jax and Sundered sky didn't lose any winrate actually. So yeah, good point but it only works to some extent with high skill expression champions.
@@armasot it differs some definitely (Azir in the same patch got scorch and gathering storm swapped, which dropped Gathering Storms winrate a solid 4% and raised scorches by 3%)
gr8 video mate love the accent lol ive hit chall in early season this year nd my goal is to get 1k lp by end of season so im just info hungry ur channel seems good will be in touch
Very good video but I want to point out something. Item win rates can be very misleading because items are very situational. For example, the mejai always have a higher win rate than other items because you buy the mejai only when you are very ahead. High win rate on an item doesn't mean it is a good item.
It works with mejais like this, yeah, because everyone knows that they should buy it only from ahead. Is there any item in league that you're also buying from ahead/behind? Not really. Well, there are some other things, like 1st item has deflated winrate because of 2nd item choice. It happened with rageblade on Kog'maw because of nashor's tooth 2nd, which is a suboptimal choice for him. Other than that, as long as sample size is big enough, winrate of the items will show you their real power.
Good video the only thing I have to add is that the winrate of items are not exactly that easy to compare when there is a large pick rate difference. An example that phreak had was that stating boots no pots on hecarim jungle one season had like a 5% higher winrate because all the players that stated with no pots were hecarim mains and won more games regardless of the starting item, according to phreak he estimated that you should assume the winrate of an off meta build will decline by about 2% if it became the meta build.
So you still have to do your own testing and conclusion of the items
Thank you for your feedback, i appreciate it!
Yeah, i agree, stats can be hard to read sometimes, i'll do a video about it someday. There are 2 important things about items winrates that you need to keep in mind.
1. is amount of games that was played with certain item. 1-2k games is the minimum to judge an item, but it can fluctuate a lot, so it's better to use larger sample size(30 days)/track an item for many patches/use gold+ stats.
2. gold cost. Low cost items should have lower winrate than higher cost items. However, when low cost item (for example statikk in Jhin's case) has more winrate than high cost item (collector), it's pretty clear which item is better.
Not sure about Phreak statement tho. I think he accounted for the margin of error in his conclusions, but the margin of error depends on the number of games played. For example, i won't even consider to buy an item with 1-2k games if it has +-the same winrate as the most popular item, because winrate in 1-2k games can fluctuate a lot.
I recently found a cool example of how pickrate doesn't mean that much so let me share it with you.
ADC Brand got some popularity not so long time ago, so he's the perfect champion to check how pickrate affected his winrate.
lolalytics.com/lol/brand/build/?lane=bottom&patch=14.8
Brand adc stats for 14.8 patch: 55.76% winrate, 9k
....a couple of patches after....
lolalytics.com/lol/brand/build/?lane=bottom&patch=30
30 days stats, 80k games, 54.92% winrate.
Worth to mention that he also got nerf between 14.8 and last 30 days, so overall, his winrate was almost +-the same, if we take into account 1 nerf.
@@armasot nah its not "margin of error" but the general pattern that players that alter their build will on average be also just better on the champ (since you generally only deviate from the common path, if you have an idea _why_ you'd deviate); this was shown with Items (Trinity vs Divine Sunderer) in S12 on a champ (which I sadly forgot); where swapping which is recommended raised the winrate of the "standard" item and lowered that of the "sleeper OP" item.
They replicated a similar effect lately with runes, where most of the time the new recommended rune lost about 2% winrate and the no longer recommended one gained about 2%.
Well, i think it depends on champion's skill expression. Champions with high skill expression will lose winrate on any recommended items and runes for about 2%, because lower elo players won't be able to play this champion no matter what (checked it with Azir change, coup de grace to cut down). For champions, who is just average in any elo this % will be much lower or won't exist at all, like with Trundle top (pta was changed to grasp as recommended, didn't lose any winrate or Brand adc, who got more popular recently, also didn't lose any winrate up until now with nerfs).
Also, with Trinity vs Divine Sunderer, i think it was Jax and Sundered sky didn't lose any winrate actually.
So yeah, good point but it only works to some extent with high skill expression champions.
@@armasot it differs some definitely (Azir in the same patch got scorch and gathering storm swapped, which dropped Gathering Storms winrate a solid 4% and raised scorches by 3%)
gr8 video mate love the accent lol ive hit chall in early season this year nd my goal is to get 1k lp by end of season so im just info hungry ur channel seems good will be in touch
Thank you, I appreciate it! Good luck with your goal
I didn’t expect such a cool channel to have so few subscribers 😢
The 5th thing is very useful. It's strange how this video has so few views
Thank you for your feedback, i appreciate it!
Rengar has R => Stand next to teammates and pray he doesn't get to click on you
Haha, Rengar is kinda scary :)
Very good video but I want to point out something. Item win rates can be very misleading because items are very situational. For example, the mejai always have a higher win rate than other items because you buy the mejai only when you are very ahead. High win rate on an item doesn't mean it is a good item.
It works with mejais like this, yeah, because everyone knows that they should buy it only from ahead. Is there any item in league that you're also buying from ahead/behind? Not really.
Well, there are some other things, like 1st item has deflated winrate because of 2nd item choice. It happened with rageblade on Kog'maw because of nashor's tooth 2nd, which is a suboptimal choice for him.
Other than that, as long as sample size is big enough, winrate of the items will show you their real power.
What rank are you?
Hey! I'm a master tier player, currently 244 lp, had 490 lp peak
@@armasot thx