According to this study you are better off squatting and deadlifting than playing on a swiss ball......
(I think I've said something similar before)
Trunk muscle activity during stability ball and free weight exercises.
The purpose of this investigation was to compare trunk muscle activity during stability ball and free weight exercises. Nine resistance-trained men participated in one testing session in which squats (SQ) and deadlifts (DL) were completed with loads of approximately 50, 70, 90, and 100% of one-repetition maximum (1RM). Isometric contractions during 3 stability ball exercises (quadruped (QP), pelvic thrust (PT), ball back extension (BE)) were also completed. During all exercises, average integrated electromyography (IEMG) from the rectus abdominus (RA), external oblique (EO), longissimus (L1) and multifidus (L5) was collected and analyzed. Results demonstrate that when expressed relative to 100% DL 1RM, muscle activity was 19.5 +/- 14.8% for L1 and 30.2 +/- 19.3% for L5 during QP, 31.4 +/- 13.4% for L1 and 37.6 +/- 12.4% for L5 during PT, and 44.2 +/- 22.8% for L1 and 45.5 +/- 21.6% for L5 during BE. IEMG of L1 during SQ and DL at 90 and 100% 1RM, and relative muscle activity of L5 during SQ and DL at 100% 1RM was significantly greater (P < or = 0.05) than in the stability ball exercises. Furthermore, relative muscle activity of L1 during DL at 50 and 70% 1RM was significantly greater than in QP and PT. No significant differences were observed in RA and EO during any of the exercises. In conclusion, activity of the trunk muscles during SQs and DLs is greater or equal to that which is produced during the stability ball exercises. It appears that stability ball exercises may not provide a sufficient stimulus for increasing muscular strength or hypertrophy; consequently, the role of stability ball exercises in strength and conditioning programs is questioned. SQs and DLs are recommended for increasing strength and hypertrophy of the back extensors.
So it is this: