Four race lengths are available: 10 laps (no damage, no yellow flags), 30 laps (no damage), 60 laps, and 200 laps. Practice and qualifying modes are also available. Practice allows car setups to be altered and tested in real time with changes taking immediate effect. Choosing not to participate in qualifying results in starting from the back of the field. The qualifying session requires four laps to be completed, with the mean of the four lap times determining the qualifying position. Car damage can occur during qualifying sessions but not during practice.
The available cars are a yellow Penske-Chevrolet, a red Lola-Buick, or a blue March-Cosworth. The Penske has the fastest default setup, though any car can compete effectively with careful tuning. The player's car is always numbered 17. Various settings can be changed during practice from menus associated with Function keys F3–F10.
A wide variety of car settings can be altered during practice. During qualifying and race sessions, all settings except turboboost and the anti-roll bars can only be altered during a pit stop, and some settings are unchangeable even then. Over-revving the engine can cause it to blow.
Six camera angles are available: In-Car, Behind, Track, TV, Sky, and Leader/Crash. The replay mode covers the previous 20 seconds of racing. The Leader/Crash camera shows crashed cars but is unavailable during practice and qualifying.
The 32 computer-controlled cars can crash at any point or retire with mechanical problems during pit stops. In a 10-lap race, a crash triggers a briefly flashing yellow flag but all cars continue at full speed. In all other race distances, yellow flags cause cars to slow and prohibit passing until the incident is cleared. A crashed car typically remains on the circuit for 2–3 laps before being cleared. Yellow flags are not displayed if the player's own car crashes, unless other cars hit the wreckage. During a yellow flag period, speeds are restricted to approximately 90 mph, against a typical race pace of up to 230 mph.
The player's car cannot be damaged in 10-lap or 30-lap races. In longer races, hard contact with a wall or another car can cause wheel or engine damage. Destruction of any two tires makes recovery from a crash extremely difficult. A large impact to the rear may cause engine damage from which there is no recovery. Possible causes of computer-car retirement include bearing, clutch, CV joint, engine, gearbox, ignition, stall, valve, vibration, radiator, and oil leak. The player-controlled car does not suffer random mechanical failures.
An Amiga version was released in late 1990, running from a single floppy disk with copy protection via a manual-based question-and-answer method. The game was identical to the MS-DOS version except in minor details. An error in programming resulted in two cars being numbered 20; in the MS-DOS version, one of these was correctly numbered 12, reflecting the actual 1989 Indianapolis 500 grid. Car control was available via mouse, joystick, or keyboard. Mouse input offered a particularly smooth and natural driving feel, with sensitivity adjustable from the main menu. One instant replay could be saved to disk, as could up to three car settings. Partly completed races could not be saved. Replays required the optional 512KB RAM upgrade.
The game was developed over 18 months with a budget of $70,000. Its theme music was produced by Rob Hubbard, who was at the time new to Electronic Arts as a music director.
The game sold more than 200,000 units by 1994. Auto racer Barry Werger praised the graphics, controls, and realism in Computer Gaming World, describing it as a "hyper-realistic simulation [and] a valuable educational tool." The Amiga version was voted the 9th best game of all time in Amiga Power. In 1994, PC Gamer UK named Indianapolis 500: The Simulation the 38th best computer game of all time, calling it "pure racing action at its best." In 1996, Computer Gaming World ranked it the 122nd-best computer game ever released.
This article is based solely on the supplied corpus. No external sources were consulted; claims that could not be substantiated against the corpus were omitted under the drop-the-claim rule.