In this paper we suggest the first systematic review and com- pare performance of most frequently used machine learning algorithms for prediction of the match winner from the teams’ drafts in DotA 2 computer game. Although previous research attempted this task with simple models, weve made several improvements in our approach aiming to take into account interactions among heroes in the draft. For that pur- pose we’ve tested the following machine learning algorithms: Naive Bayes classifier, Logistic Regression and Gradient Boosted Decision Trees. We also introduced Factorization Machines for that task and got our best re- sults from them. Besides that, we found that model’s prediction accuracy depends on skill level of the players. We’ve prepared publicly available dataset which takes into account shortcomings of data used in previous research and can be used further for algorithms development, testing and benchmarking.
We use eSports data to construct an empirical model to measure the effect of diversity on team performance. Different kinds of diversities are considered, diversity of culture, diversity of language and diversity of skill. Our main results are that cultural diversity is beneficial for team performance: the absence of diversity reduces performance by 30%. However, language and experience diversity negatively affect results. Taking the difference in the results into account, we conclude that firms should not thoughtlessly maximize team diversity: different kinds of diversity have different integration and communication costs.
This study provides readers with new information about key drivers of performance in the emerging area of eSports. Competitive computer gaming (eSports) is becoming increasingly popular, and the number of gamers and amount of prize money is growing. We therefore explore some key country-level characteristics that may contribute to players’ success, measured as money won. We use gamers’ prize earnings aggregated by country and a hurdle model to understand the determinants of performance. The results show that a 1% increase in GDP per capita leads to a 2.2% increase in prize money per capita. Country population is not statistically significant in the outcome model. This finding may indicate that eSports talents are not uniformly distributed across the world population. Surprisingly, post-Soviet and planned or post-planned economies are more likely to participate in eSports.
Smoking is a problem, bringing signifi cant social and economic costs to Russiansociety. However, ratifi cation of the World health organization Framework conventionon tobacco control makes it possible to improve Russian legislation accordingto the international standards. So, I describe some measures that should be taken bythe Russian authorities in the nearest future, and I examine their effi ciency. By studyingthe international evidence I analyze the impact of the smoke-free areas, advertisementand sponsorship bans, tax increases, etc. on the prevalence of smoking, cigaretteconsumption and some other indicators. I also investigate the obstacles confrontingthe Russian authorities when they introduce new policy measures and the public attitudetowards these measures. I conclude that there is a number of easy-to-implementanti-smoking activities that need no fi nancial resources but only a political will.
One of the most important indicators of company's success is the increase of its value. The article investigates traditional methods of company's value assessment and the evidence that the application of these methods is incorrect in the new stage of economy. So it is necessary to create a new method of valuation based on the new main sources of company's success that is its intellectual capital.