[DATA] Item usage in 6.71 (incomplete)

Gheizen64 · 03-05-2011, 11:15 PM

Yes, i skipped 6.70, and yes, i'm only showing certain items instead of them all as usual. 6.70 was basically a 1 month release and was almost never played since every tournament was either 6.69 or 6.71. As far as 6.71 go, i've collected all the data i'll use but i'm analyzing it only now since the last two months were exam months and i had also other issues. For now you can look at this data and be sure i'll add most of the other items soon:

Quote:

Data from 60 games

Power Treads
- 398 total, 6.633 apg

Phase Boots
- 77 total, 1.283 apg

Arcane Boots
- 30 total, 0.500 apg

Boots of Travel
- 34 total, 0.567 apg

Bracers
- 62 total, 1.033 apg

Wraiths
- 26 total, 0.433 apg

Nulls
- 31 total, 0.517 apg

Magic Wand
- 313 total, 5.217 apg

Magic Stick
- 41 total, 0.683 apg

Poor Man's Shield
- 5 total, 0.083 apg

Ring of Basilus
- 78 total, 1.300 apg

Urn of Shadows
- 40 total, 0.667 apg

Janggo
- 29 total, 0.483 apg

Soul Ring
- 22 total, 0.367 apg

Medallion of courage
- 29 total, 0.483 apg

Hand of Midas
- 0 total, 0.000 apg

Force Staff
- 58 total, 0.967 apg

Dagger
- 105 total, 1.750 apg

Vanguard
- 86 total, 1.433 apg

Planeswalker's Cloack
- 42 total, 0.700 apg

Hood of Defiance
- 31 total, 0.517 apg

Pipe
- 36 total, 0.600 apg

Headdress of Rejuvenation
- 10 total, 0.167 apg

Mekansm
- 65 total, 1.083 apg

Black King Bar
- 60 total, 1.000 apg

Manta Style
- 40 total, 0.667 apg

Sange and Yasha
- 14 total, 0.233 apg

Lothar's Edge
- 11 total, 0.183 apg

Butterfly
- 33 total, 0.550 apg

Ethereal Blade
- 9 total, 0.150 apg

Diffusal Blade
- 7 total, 0.117 apg

Helm of the Dominator
- 14 total, 0.233 apg

Stygian Desolator
- 3 total, 0.050 apg

Maelstrom
- 3 total, 0.050 apg

Mjollnir
- 3 total, 0.050 apg

Radiance
- 25 total, 0.417 apg

Monkey King Bar
- 15 total, 0.250 apg

Guinsoo
- 35 total, 0.583 apg

Aghanim Scepters
- 35 total, 0.583 apg

Orchid's Malevolence
- 10 total, 0.117 apg

Shiva's Guard
- 14 total, 0.233 apg

Linken's Sphere
- 32 total, 0.533 apg

Blademail
- 17 total, 0.283 apg

Assault Cuirass
- 11 total, 0.183 apg

Heart of Tarrasque
- 10 total, 0.167 apg

Ghost Scepter
- 33 total, 0.550 apg

PT took a massive hit, people have finally started to feel that 10->8 nerf. From 7.5 to 6.6, and AB (0,3->0,5), Bot and PB (1,1->1,3) all went up slightly due to this. Vanguard is also decreased slightly, mostly because of Janggo+S&Y being some decent alternative for early hp. Other items soon to come. Manta going down a lot, S&Y going slightly up.

EDIT1: added Urn, Soul Ring, Bkb, Medallion and Janggo numbers
EDIT2: added force staff and dagger
EDIT3: added cloack, headdress, hood, pipe, meka, agha
EDIT4: added butterfly, EB, ghost scepter, Diffusal, Hotd, MKB, Lothar, Deso, Orchid, Mael, Mjoll

CaptainPlatypus · 03-06-2011, 12:06 AM

Thanks Gheizen! Sorry if I nagged you in your suggestion thread.

DemonDays · 03-06-2011, 12:09 AM

so wand will slowly be replaced by urn or medallion?

Gheizen64 · 03-06-2011, 12:28 AM

Quote:

Originally Posted by CaptainPlatypus

Thanks Gheizen! Sorry if I nagged you in your suggestion thread.

Np, you did good remembering me.

CaptainPlatypus · 03-06-2011, 12:36 AM

From what we have, holy crap PTs are going out of fashion fast. Bracers have seriously fallen off too (1.5 to 1 per game), which doesn't make much sense, especially with the introduction of janggo - maybe people have found something better to do on their supports than build cheap hit points. Manta hit too - yeesh. Guess the decreased ranged image damage is finally starting to get noticed. Even basilius went down a little. I'm guessing we'll see the new items picking up a fair amount of the missing share, and the rest going to things like mek and hood. Not sure if the missing mantas are getting replaced by anything in particular - SnY's rise is starting, but not a terribly big deal yet. Probably just a bunch of minor rises in every other carry-type item.

Quote:

Originally Posted by DemonDays

so wand will slowly be replaced by urn or medallion?

I'm actually looking forward to seeing urn data - it and wand both peaked pretty hard in 6.68 before falling off in 6.69, so I want to see if, like wand, it's continued its decline, or if it's gone back up. Anecdotal evidence (my memory of watching replays) suggests the latter, but I want to know for sure. I also want to see medal/janggo use (though the first version they were competitively used in will probably not be very indicative of their long term usage level), BKB (it was stable in 6.69, but gained use in every other version since gheizen started in 6.66), dagger/force staff (dagger fluctuations in particular can be used to watch trends in the prevalence of AoE initiation - at least, that's my theory), and guinsoo (saw a BIG jump in 6.69).

EDIT: actually, is there any way to track consumables as well? Be interesting to see how many times smoke is bought, for example - and observers and sentries couldn't hurt either. I'm not sure how you gather the data, though, so I have no idea if this would be practical.

Gheizen64 · 03-06-2011, 09:01 AM

Quote:

Originally Posted by CaptainPlatypus

From what we have, holy crap PTs are going out of fashion fast. Bracers have seriously fallen off too (1.5 to 1 per game), which doesn't make much sense, especially with the introduction of janggo - maybe people have found something better to do on their supports than build cheap hit points. Manta hit too - yeesh. Guess the decreased ranged image damage is finally starting to get noticed. Even basilius went down a little. I'm guessing we'll see the new items picking up a fair amount of the missing share, and the rest going to things like mek and hood. Not sure if the missing mantas are getting replaced by anything in particular - SnY's rise is starting, but not a terribly big deal yet. Probably just a bunch of minor rises in every other carry-type item.

I'm actually looking forward to seeing urn data - it and wand both peaked pretty hard in 6.68 before falling off in 6.69, so I want to see if, like wand, it's continued its decline, or if it's gone back up. Anecdotal evidence (my memory of watching replays) suggests the latter, but I want to know for sure. I also want to see medal/janggo use (though the first version they were competitively used in will probably not be very indicative of their long term usage level), BKB (it was stable in 6.69, but gained use in every other version since gheizen started in 6.66), dagger/force staff (dagger fluctuations in particular can be used to watch trends in the prevalence of AoE initiation - at least, that's my theory), and guinsoo (saw a BIG jump in 6.69).

EDIT: actually, is there any way to track consumables as well? Be interesting to see how many times smoke is bought, for example - and observers and sentries couldn't hurt either. I'm not sure how you gather the data, though, so I have no idea if this would be practical.

Consumables are out of the question, i register the data via screenshot of the final score table. Bracers are going down, but i suspect a lot of the "missing" bracers are in fact Janggos. I can already say which items are still unused:

midas
bloodstone
mael-mjoll
basher
armlet

Nothing new to see here.

laikzasve · 03-06-2011, 10:20 AM

At any rate gj,and thank you for effort

CaptainPlatypus · 03-06-2011, 11:34 AM

Quote:

Originally Posted by Gheizen64

Consumables are out of the question, i register the data via screenshot of the final score table. Bracers are going down, but i suspect a lot of the "missing" bracers are in fact Janggos. I can already say which items are still unused:

midas
bloodstone
mael-mjoll
basher
armlet

Nothing new to see here.

Ah, gotcha. That explains the bracers quite nicely. It might be worthwhile to consider switching your data method - perhaps make a mouse/keyboard macro (with autohotkey or similar) that automatically strips item data from DotAReplay (since DotAReplay shows all purchased items)? With the built-in "export to plain text" and "detailed report (item build)" functions, it shouldn't be that hard to write, and I could try helping out if it would be useful. An example of the plaintext you could write a macro to automatically export (from NaVi vs Ends, because I haven't deleted it yet) is below:

From there, all you need is a way to import that into whatever database you have, and from plaintext that shouldn't be that difficult. This may not be something you're interested in or have the time/patience for - the current version is still invaluable - but I thought I'd suggest it.

Unstable · 03-06-2011, 11:37 AM

It's good some one actually does something for this pretty much shitty community while asking for nothing in return. Good job with the statistics and keep up the good work.

elsoloist · 03-06-2011, 11:53 AM

I must agree with unstable. it's hard to continue doing this tabulation of data version after version, showing us some things that might not be as intuitive as we thought, that have been noticed by the better players.

Gheizen64 · 03-06-2011, 01:15 PM

Woah, bkb almost disappeared (1.7->1). BkB-less carries are probably getting more popular this version, things like Spec, Razor, Weaver, Dusa etc...

CaptainPlatypus · 03-06-2011, 01:32 PM

Quote:

Originally Posted by Gheizen64

Woah, bkb almost disappeared (1.7->1). BkB-less carries are probably getting more popular this version, things like Spec, Razor, Weaver, Dusa etc...

Holy. That's a damn big change. And yeah, razor/spec/weaver probably have a lot to do with it, as well as the general lower-carry atmosphere.

elsoloist · 03-07-2011, 12:31 AM

Since janggo is out, we can safely say that there's no change in bracer buying rate, considering if u add the 2 rates up, u'll get exactly 1.5, which is the old buyrate for bracer..

CaptainPlatypus · 03-07-2011, 01:45 AM

Dagger fell off too, holy crap. Weird version. :P

drafter · 03-07-2011, 02:34 AM

Quote:

Originally Posted by CaptainPlatypus

[Use a replay parser]

Replay parsers forget about some items sometimes - note how 818 gets only one GG branch before the Magic Wand recipe in your example. If there is a way to work around this it would allow a much larger dataset to be obtained really fast, without the potential to human error and in a way that can take into account items that are sold or used in recipes.

CaptainPlatypus · 03-07-2011, 03:24 AM

Quote:

Originally Posted by drafter

Replay parsers forget about some items sometimes - note how 818 gets only one GG branch before the Magic Wand recipe in your example. If there is a way to work around this it would allow a much larger dataset to be obtained really fast, without the potential to human error and in a way that can take into account items that are sold or used in recipes.

IIRC, this glitch only happens when the same player buys the same item very rapidly - meaning almost always with missing branches, which isn't the biggest deal in the world. I could be wrong, though.

qoou · 03-07-2011, 03:29 AM

Quote:

Originally Posted by drafter

Replay parsers forget about some items sometimes - note how 818 gets only one GG branch before the Magic Wand recipe in your example. If there is a way to work around this it would allow a much larger dataset to be obtained really fast, without the potential to human error and in a way that can take into account items that are sold or used in recipes.

Erm. It says he gets three.

LiquidNitrogen · 03-07-2011, 04:12 AM

you go forcestaff

<3

Sonicspear · 03-07-2011, 04:29 AM

Thanks Gheizen. I forgot, but didn't someone also do the amount of heroes picked/banned per tournament? Maybe having those figures alongside the item figures will help get a clearer picture.

Could be the usage of highly mobile characters like Weaver and Spectre or characters who have a long range AoE target spell like AA that could be factors into the decrease in Blink/BKB.

CaptainPlatypus · 03-07-2011, 05:18 AM

Quote:

Originally Posted by Sonicspear

Thanks Gheizen. I forgot, but didn't someone also do the amount of heroes picked/banned per tournament? Maybe having those figures alongside the item figures will help get a clearer picture.

Could be the usage of highly mobile characters like Weaver and Spectre or characters who have a long range AoE target spell like AA that could be factors into the decrease in Blink/BKB.

That much I'd be willing to go through and do, assuming all we want is hero ban/pick rate and win/loss - I can spend 15 minutes on gg.net and 15 more formatting as well as anybody else. It'll be manual. Because I'm old school.

03-06-2011, 12:06 AM	#2
CaptainPlatypus Member Join Date: Aug 2009 Posts: 939 Blog Entries: 1	Re: [DATA] Item usage in 6.71 (incomplete) Thanks Gheizen! Sorry if I nagged you in your suggestion thread. __________________ Please Stuff Counterstompin' Think (and search) before you post. Strategy+ My item suggestion, the Illusory Manteau - check it out! My (draft, still very incomplete) Invoker guide with Val

03-06-2011, 12:09 AM	#3
DemonDays Member Join Date: Sep 2010 Posts: 1,034	Re: [DATA] Item usage in 6.71 (incomplete) so wand will slowly be replaced by urn or medallion?

03-06-2011, 10:20 AM	#7
laikzasve Member Join Date: Oct 2009 Location: Serbia Posts: 1,545	Re: [DATA] Item usage in 6.71 (incomplete) At any rate gj,and thank you for effort __________________ Ok,new suggestions are ready.Feel free to criticize them : Tiny-like combo for magnus!! Simple solution to make Shredder easier to gank,while making him funnier to play. Nerf to Dragonus Rebalance of Concussive Shot

03-06-2011, 11:37 AM	#9
Unstable Member Join Date: Oct 2009 Location: The middle of nowhere - Bulgaria Posts: 815	Re: [DATA] Item usage in 6.71 (incomplete) It's good some one actually does something for this pretty much shitty community while asking for nothing in return. Good job with the statistics and keep up the good work. __________________ *Nobody is perfect. I am nobody. I am perfect.*

03-06-2011, 11:53 AM	#10
elsoloist Member Join Date: Jun 2009 Posts: 406	Re: [DATA] Item usage in 6.71 (incomplete) I must agree with unstable. it's hard to continue doing this tabulation of data version after version, showing us some things that might not be as intuitive as we thought, that have been noticed by the better players.

03-06-2011, 01:15 PM	#11
Gheizen64 Member Join Date: Jun 2009 Posts: 2,216	Re: [DATA] Item usage in 6.71 (incomplete) Woah, bkb almost disappeared (1.7->1). BkB-less carries are probably getting more popular this version, things like Spec, Razor, Weaver, Dusa etc... __________________ Analysis of item usage in competitive 6.66, 6.67, 6.68, 6.69 and 6.71. Ongoing. Look my reorganization of the weapon shop for: • Organized shop layout (ala Sena the Accesorizer) • More unique +dmg items (BoA ,Broadswoard & Claymore) • Items and recipes are more logical and intuitive • Hammer is the old Javelin (the %chance fits naturaly in Basher + Mael) • Desolator is now made from swords and more!

03-07-2011, 12:31 AM	#13
elsoloist Member Join Date: Jun 2009 Posts: 406	Re: [DATA] Item usage in 6.71 (incomplete) Since janggo is out, we can safely say that there's no change in bracer buying rate, considering if u add the 2 rates up, u'll get exactly 1.5, which is the old buyrate for bracer..

03-07-2011, 01:45 AM	#14
CaptainPlatypus Member Join Date: Aug 2009 Posts: 939 Blog Entries: 1	Re: [DATA] Item usage in 6.71 (incomplete) Dagger fell off too, holy crap. Weird version. :P __________________ Please Stuff Counterstompin' Think (and search) before you post. Strategy+ My item suggestion, the Illusory Manteau - check it out! My (draft, still very incomplete) Invoker guide with Val

03-07-2011, 04:12 AM	#18
LiquidNitrogen Member Join Date: May 2010 Location: Seattle, WA Posts: 2,854	Re: [DATA] Item usage in 6.71 (incomplete) you go forcestaff <3 __________________ David Park Photography Like it motherfuckers!