[Aug] TLPD Race Winrate Graphs - Page 6

On September 05 2011 22:09 Truedot wrote:


looking back at this graph I;ll give you an example.

http://i.imgur.com/bdP2e.png

See games in may for all races. The lowest win rate and the highest win rate oppose each other.

They add up to 100 together which converts to 100%.

Example 1: PvT may
45.1 vs 55.9 opposing numbers in percentages, add them together to get 100.

Example 2:
ZvT may.
59.3 vs 40.7

again, the numbers oppose each other. add them up to 100.

this is basic math.

On September 05 2011 22:14 Truedot wrote:



Because MC and other people changed how people play normally to almost always blink stalker. Its more effective than a ball because of its ridiculous mobility.

zerg status mobility with protoss damage and HP values, and the ability to have ever living units, vs building a ball that gets chain fungaled.

Its still viable, its still damn deadly in the right hands, and it still overpowers a zerg that doesn't just mass infestors forever, but its not as efficient as blink stalker.

On September 05 2011 20:16 Hondelul wrote:
Update:
Since all percentages from July changed from [July] to [Aug] I´m quite sure the stats got combined. No seperation for International and Korea.
If I´m wrong please tell me!

On September 05 2011 22:32 ladyumbra wrote:
Forgive me as I've only been following for a few months but what happened back in January to cause zerg to have a sudden plummet to 42% when both toss and Terran was 50? Also things seemed stable for all three in June but now we are seeing a bit of disparity again as everyone does the think up new strats dance.

On September 05 2011 22:24 SwampZero wrote:



is it possible that youre really this stupid? Is there a universe out there in which such a stupid person exists?

On September 05 2011 22:23 MrCon wrote:
For the P who missed this weekend games (code A and TLOpen), I really recommend watching the protoss runs in those.
In the TLOpen I saw more warp prism in 6 hours than in 1 year, and used in a non gimmicky way.
There was some really good games. I saw threads in the strat forum about those builds already.

On September 05 2011 22:30 Ctuchik wrote:

It's both Korea and International.

There is no "combination" of the stats, that is, it's not cumulative. The reason the July numbers have changed is that I reload all the data from TLPD every month, so any games added later will now show up. Sometimes tournaments are late with submitting their results, or the TLPD guys are busy.

On September 05 2011 22:49 MrCon wrote:

Your statement is confusing. Both international and korea games are just represented now in the graph, right ?

On September 05 2011 22:54 Ctuchik wrote:


Hrrm, yeah sorry. That was confusing. =P

It is both Korea and International combined. Just want to make it clear that the data is not cumulative!

On September 05 2011 22:45 chestnutcc wrote:
A v subjective and amateur viewpoint here; apologies beforehand. I used to play age of empires and age of mythology extensively a few years back, and could not help but notice some interesting distinctions between that series and starcraft 2 (I have no experience with broodwar).

In the age series, macro focused around four resources and the ease of collection of each varied according to its relative importance. For example, gold (or stone) would be randomly flung about the map in concentrated amounts, wood was abundant for the most part etc. Additionally, resource collection required only a low cost dumping structure (such as a granary)*. The effect of this multiplicity of resources was that harassment (or raiding as it is known in that community) was a v integral part of the game. Constant raiding was a feature of even the most macro oriented games (the age series had a interesting blend of hard macro and intensive harass). Since games did not usually end with a few decisive encounters, it was the ability of players to resupply and macro up constantly that won games, and this was what the continuous harass targeted.

iamke55 in an excellent thread** outlines the PvT metagame, which ends in the current state (forget the 1-1-1 for now) of terran taxing protoss multitasking with constant harass. Terran in particular are v suited to this style of play, all thanks (imo) to the mule. This is critical to their playstyle. Count the raiding avenues they have: Reapers, hellions, banshees and drops.

The shift in the zerg meta game may be attributed to the aggressive nestea-losira style (in tastosis terms, nestea being pure reactionary harass, while losira forcing his will on the opponent with constant harass). Zerg have speedlings, mutas and even baneling drops (nydus worms are less seen).

Toss have at best, blink stalkers and phoenix, the latter of which I think are neutralized by a few turrets. I am unsure of the efficacy of DTs, they seem to do better late game, when the opponents attention is stretched over multiple bases, early to mid game a single detector can neutralize them, and they are (taking the tech path into consideration) cost ineffective. Warp prisms are rarely seen, and early to midgame appear to be cost ineffective, due to the constraints the macro game places on toss army size (all ins aside). The interesting thing is that both terran and zerg can effectively pressure from the get go, toss is at a disadvantage here; this is sort of how the game is supposed to be. It would seem toss needs a raiding unit of some sort, to exert some counter pressure cost effectively***.

*This basically meant a single raid wasn't usually game ending, good macro could always resupply efficiently. It was only one part of a larger war of attrition, and effects multiplied the longer the game got. Also note that four resources provided more avenues for effective harass; in sc2 the concentration of two resources at one generally hard to access and defensible spot means the efficacy of harass, when it happens, is vastly increased i.e. a single good raid can end the game.The distinction between harassment and timing attacks is obvious.
**http://www.teamliquid.net/forum/viewmessage.php?topic_id=196385
***The inability to do this imo, provides timing windows, which builds like the 1-1-1 exploit.