Simulator, ball machine, or hitting partner? what each one actually trains
By: JoeMichael Olfindo, Tennis Director and Partner
I get asked which one is best. However, in reality, they train different things, and the players who improve fastest use all three on purpose.
Here’s the honest breakdown, including where the simulator we run loses.
What a hitting partner gives you that nothing else can
Start here, because I don’t want to oversell technology.
A live opponent produces something no machine does: disguise and consequence. You have to read a body before the ball is struck. You have to decide where to hit it under the knowledge that a person is going to punish a bad choice. You experience the pressure of a scoreboard.
Those are real skills, they’re trained nowhere else, and they’re why a player who only drills gets beaten by a player who only plays.
What a partner is bad at: volume and consistency. In a typical hour of social doubles you might make clean contact with 120 to 200 balls, and every one of them arrives differently. That’s great for decision-making and terrible for building a stroke, because a stroke is built by repetition against a stable input. Partners are also unreliable, expensive if you’re paying a hitter, and — in Houston, most months of the year — subject to weather.
Use hitting partners for: point construction, pressure, tactics, disguise. Once a week minimum if you compete.
What a ball machine is genuinely good at
A ball machine does one thing extremely well: it delivers a huge number of near-identical balls - albeit almost always dead balls.
If you’re rebuilding a forehand, you need hundreds of repetitions with the same input so you can isolate one variable. A machine gives you that. It also never gets tired, never judges you, and doesn’t mind that you’re working on something ugly.
What a machine is bad at: everything after contact. A machine has no idea whether your ball went in. Feed a machine two hundred balls and you can hit two hundred beautiful shots four feet long and never know. It also can’t vary intelligently, and it gives you no feedback loop, which means bad habits get reinforced with the same efficiency as good ones. That’s the real risk — a machine amplifies whatever you’re already doing.
Use it for: stroke mechanics, grooving a change outdoors.
What a simulator adds
The reason we built around simulators is a specific gap: repetition volume with outcome feedback.
Our bays use real tennis balls with motion-tracking sensors, so every shot you hit is measured — speed, distance, and where it actually landed. You hit a few hundred balls in 25 minutes, which is machine-level volume, and each one tells you whether it worked.
That closes the loop a ball machine leaves open. You’re not guessing whether that adjustment helped. You can see spin go up or placement tighten across a session, and it turns out that measurement changes behavior — players work harder and more precisely when there’s a number attached.
The target work matters too. Free hitting produces vague intentions. Aiming at a defined target produces a specific one, and specific intention is most of what separates practice that transfers from practice that doesn’t.
Also - you keep hitting. No spending 10 minutes between sets picking up balls all over a court like after a ball machine session. The simulator recirculates the balls, so you never run out.
What a simulator is bad at, honestly:
It’s not a court. You’re not covering a full court, not chasing wide balls, not learning to recover to the center after a stretched forehand. Movement over distance is trained on a court, period. That’s exactly why we added two full courts alongside more simulators this month.
No disguise. The feed doesn’t lie to you or hide its intentions. Reading an opponent is a court skill.
No opponent psychology. Nobody in a bay is going to change tactics because they noticed your backhand is shaky.
Doubles doesn’t exist. Positioning, partner movement, poaching — court only.
Anyone who tells you a simulator replaces court time is selling you something. It replaces the unavailability of court time, doesn’t cancel on you or get sick, and doesn’t get wet with the weather. It’s reliable again and again to keep you training and improving.
How I’d build a week
Improving recreational player, 3 sessions/week: at least one simulator session on mechanics, one class, and one match. The simulator session is where you change or work on something specific; the match is where you find out if it took.
Competitive adult, 4–5 sessions: two simulator sessions, one class, one or two matches. Add a private lesson monthly to make sure the thing you’re grooving is the right thing.
Junior, developing: majority court and live-ball, with simulator sessions used surgically — reps, a specific stroke change, and extra volume, especially on week’s where weather killed a court day (or three).
Adult beginner: front-load the simulator. Volume of clean contact is the entire game in your first three months, and a simulator gives you five times the reps of a group class without an audience.
The real answer
The best training tool is the one you’ll use twice a week for a year.
That’s the unglamorous truth. A perfect program you do for three weeks loses to a decent program you keep. Most players’ constraint isn’t which tool is optimal — it’s that their tennis depends on a court being free, a partner showing up, and it not being 100 degrees or rainy.
Take those variables out and consistency becomes possible. Then use the court time you do get for what only a court can teach.
