While waiting for the Dueling Fates update, the Dota 2 community keeps creating content of any kind. The Reddit user DotA2Analyst shared a statistical article on heroes’ shoes, number of feet, and their impact on pick and win rates. We asked his permission to publish it here.
"Shoes in DotA2 are a poorly studied topic. To test these uncharted waters, this study will seek to define pertinent demographics of DotA2 hero shoes.
At this point, it is important to stress that I do not mean the purchasable boots in game, but rather the shoes that heroes themselves may or may not be wearing in their base models.
Additionally, for the purposes of this study, shoes were defined as fully closed-toe (example here); sandals, wraps, or any other piece of footwear which exposed skin from the ankle down were not considered “shoes” (example here).
First, it is necessary to break down heroes based on number of feet. The table below compares the three categories of shoe-wearing used in this study (wears shoes, does not wear shoes, has no feet to wear shoes) to the number of hero feet.
As seen above in Table 1, there is no significant difference in whether heroes with 2 feet wear shoes or do not (wear shoes n=38 does not wear shoes n=44).
However, heroes with at least 4 feet (n=12 p<0.01) exclusively do not wear shoes, suggesting a strong negative correlation between quadruped (4 legs)/ insectoid (6 legs) leg distribution and choice to not wear shoes.
To further study hero shoe demographics, the three shoe-wearing categories were next compared to gender.
Table 2 highlights interesting trends: male heroes show a moderate to strong tendency to not wear shoes (p<0.05), while females slightly lean towards a shoe wearing preference.
Of note, not a single agender hero wears shoes (p<0.001) which could possibly be attributed to the fact they all fly or float (Io, Jakiro, Phoenix, Puck, Visage).
Shoe demographics were then studied in relation to hero faction (Radiant or Dire) in table 3 above. Radiant heroes were relatively split between wearing (n=25) and not wearing shoes (n=29).
Interestingly, there is a significant difference (p<0.01) between shoe status among Dire heroes, with most going shoeless (n=27 no shoes and n= 13 with shoes). This finding is surprising, given the harsher landscape and habitat on the Dire side of the map (fires, hard and rocky ground) compared to the plush grass of the Radiant side.
Next, heroes were separated into their three attributes (strength, intelligence, and agility).
There are two main findings of interest when comparing attribute to shoe-wearing.
First, intelligence based heroes show a strong tendency towards not wearing shoes (no shoes n=23, shoes n= 10 p<0.01). The shoes may get in their way of arcane and magic abilities, which are generally more important to intelligence heroes.
Second, while shoe status is relatively split among agility heroes, there are a shockingly high number of heroes (n=9) that have no feet—the highest of the three attributes—which is strange. To have a surprisingly large proportion of heroes without feet in an attribute based on agility goes against logic.
The table below highlights the three shoe-wearing categories, and the average win rates, and average number of times picked for all heroes in those groups.
While there is no significant difference between win rates of shoe wearing and non-shoe wearing heroes, heroes without feet have a win rate more than 1% higher than those two groups (p<0.01)!
Considering feetless heroes are picked the least, these data suggest a potential important consideration to make during drafting. There are 19 heroes without feet, and statistically speaking, picking one of those 19 will give you a better chance to win the game than picking heroes with feet—no matter their shoe wearing preference.
In conclusion, this pilot study has presented data on hero shoe demographics previously not reported in literature. There are significant differences among hero leg number, gender, attribute, and faction when it comes to shoe wearing.
Additionally, players picking habits show preferences towards heroes with feet who wear shoes, but this preference is perhaps misguided, as heroes without any feet perform better than their footed counterparts."